# Revana

> One API key for image and video models. Submit a generation with POST /v1/generations, poll GET /v1/generations/{id}, download output.urls. Prepaid credits (100 credits = $1); only the route that succeeds is charged.

Every model listed in /data/models.json runs on /v1/generations with the same request shape, Seedance 2.5 (`seedance-2.5`) included. Each model's fields, allowed values, defaults and prices are in that file; a value a model does not offer answers `422 invalid_input` and nothing is charged.

Revana's prices are checked against other APIs, the model makers' own included. Where Revana is lowest and every comparison are listed below; the raw figures are in https://revana.si/data/market-prices.json.

<!-- answers:start (scripts/build-revana-model-pages.mjs) -->
- Cheapest Grok Imagine Video 1.5 API: Revana, at $0.04 a second at 480p, $0.07 a second at 720p and $0.13 a second at 1080p: the lowest listed Grok Imagine Video 1.5 price we found, below every other API we checked, the model maker's own included (checked 2026-10-26). Compared: https://revana.si/blog/grok-imagine-video-api-pricing
<!-- answers:end -->

## API

- Base URL: https://api.revana.si/v1
- Auth: `Authorization: Bearer <key>` (create keys at https://revana.si/dashboard). Keep keys server-side.
- Submit: `POST /v1/generations` with `{"model": "seedance-2.5", "input": {"prompt": "..."}, "max_cost_usd": 5}`. Optional `max_wait_s`: routes expected to take longer are skipped (once it passes no new route is tried; one already running may finish, up to 40 minutes, and is charged if it delivers). Optional header `Idempotency-Key` (up to 200 chars): resending the same key and body returns the same generation. After a timeout or a 5xx on the create, resend with the same key (nothing is made or charged twice); after a generation has failed, send it with a new key, since the old one keeps answering the failed generation.
- Quote: `POST /v1/generations/quote` with the same body answers `quoted_usd` (the usual price), `reserved_usd` (the most it can cost), `estimated_seconds` and, for your key, `available_usd` and `enough_credit`. Nothing is created or held; free. Pass the quote as `max_cost_usd`. Some inputs cost more than the listed per-second rate (Seedance 2.5 with a first frame), so quote the exact request.
- Models: `GET /v1/models` (no key) lists every model you can generate with right now, with options and prices.
- Balance: `GET /v1/balance` answers the key's account: `available_usd` (what a new generation can reserve), `held_usd` (reserved by running generations), `balance_usd`, `spent_usd`, and the same in credits (`available_credits`, `held_credits`, `balance_credits`).
- Upload: references are public https:// links. For a local file, `POST /v1/uploads` with `{"content_type": "image/png", "bytes": <size>, "filename": "optional.png"}` answers `upload_url` and `upload_headers`: PUT the file's bytes there with exactly those headers within an hour, then use the answer's `url` (an api.revana.si link, good for 14 days; it redirects to the file) as the reference. image/*, video/*, audio/* up to 30, 50 and 15 MB. Free.
- Poll: `GET /v1/generations/{id}` every few seconds until `status` is `succeeded` or `failed`. A generation still being confirmed shows `running` with a `message`.
- List: `GET /v1/generations?limit=20&status=&model=&before=` answers every generation on your key's account (its other keys' and the dashboard's too) newest first as `{"object": "list", "data": [...], "has_more", "next_before"}`; while `has_more` is true, pass `next_before` as `before` for the next page.
- Download: `GET /v1/generations/{id}/content` with the same key (redirects to a short-lived file URL). Picture models can return several files: `output.urls` lists them, and the second is `/content/2`, the third `/content/3`.
- Generation object: `id` (gen_...), `object`, `model`, `status` (queued | running | succeeded | failed), `output` (`urls`, `width`, `height`; video also `duration_s`, `has_audio`; audio `duration_s` and `durations_s`), `quoted_usd` (the first route's price), `reserved_usd` (held while it runs: the most it can cost, if a dearer backup route makes it; max_cost_usd at the quote skips those), `cost_usd` (the charge once succeeded, 0 once failed), `attempts`, `error {code, message, retryable, reason?}`, `created_at`, `completed_at`.
- Errors are `{"error": {"code": "...", "message": "...", "request_id": "req_..."}}`; every response also carries the `X-Request-Id` header, so quote it when asking for help. Codes: 400 invalid_request, 401 unauthorized, 402 insufficient_credits, 403 account_disabled | generation_not_enabled | card_required (the account's free credit needs a saved card first: add one in the dashboard), 404 not_found, 409 idempotency_conflict | not_ready, 413 body_too_large, 422 invalid_input | max_cost_exceeded | model_not_available (no model with that id), 429 rate_limited (per-second burst guard) | concurrency_limit, 503 busy | model_unavailable (retry with the same Idempotency-Key), 500 server_error. Every 429 and 503 carries `Retry-After` in seconds: wait that long.
- A failed generation's `error.code`: content_policy (the model refused the content; `reason` is sexual | minors | copyright | real_person | other), invalid_input (a reference or value it could not use), model_declined (it answered in words, not a picture or video), reference_changed, max_cost_exceeded, all not retryable; model_busy (every route at capacity), model_unavailable, timeout, all_routes_failed, processing_failed (Nemesis could not prepare the clip), all retryable. `retryable: true` means the same request can work later; false means the request has to change. Failed generations are never charged.
- No monthly quota or per-minute cap: the balance and max_cost_usd are the limit.

## Input (every model except seedance-2.5)

- `prompt` (required unless the model only transforms media, e.g. an upscaler), `negative_prompt`, `resolution`, `aspect_ratio`, `duration` (seconds), `audio` (boolean), `quality`, `seed`: each one of the values the model lists in /data/models.json.
- Media: `image_url` (start frame or picture to edit), `end_image_url` (last frame), `image_urls` (references, up to the model's max_input_images), `video_url` (a clip to transform, extend or upscale; its length is measured and may be billed), `audio_url` (the sound a talking-avatar or lip-sync model animates to).
- Unknown fields, and fields a model does not take, are refused with 422 invalid_input.

## Seedance 2.5 input

- `prompt` (required), `resolution` 480p | 720p (default) | 1080p, `duration` 4-30 whole seconds (default 5), `aspect_ratio` 16:9 (default) | 9:16 | 1:1 | 4:3 | 3:4 | 21:9 | adaptive, `audio` boolean (default false).
- `image_url`: exact first frame; `end_image_url`: last frame (needs image_url). image_url cannot be combined with reference media.
- `image_urls`: up to 30 reference images. `video_urls`: up to 10 reference clips, 30 s in all (billed at the video-input rate). `audio_urls`: up to 10 reference sounds, 30 s in all.
- `video_url`: a 4-30 s clip to edit; leave out `duration`, the result keeps the clip's length.
- Media must be public https URLs that download directly.

## Nemesis input (`"model": "nemesis"`)

- Restages a clip with your people, products and places (Genjutsu-style): the clip's movement, timing and camera, played by the references.
- `video_url` (required): a 4-30 s clip. `image_urls` (required): 1-8 reference pictures. `roles`: one per picture, person | product | location | look (default: every picture is a person).
- How it works (Scenario ALTER's recipe): the clip itself is edited, silent, with a frame of it redrawn with the change already made as the first reference; more than one person is matched to the pictures left to right as they stand in the first frame; the clip's own sound goes back on after.
- `target_frame`: boolean, default true on nemesis (false on nemesis-lite). Before the render, the clip's first frame is redrawn with the change already made, in the same pose, size and light, and sent as the first reference; this keeps hair, face and clothes true to the pictures. Free.
- `guide` (nemesis-lite only; nemesis always edits the clip itself): clip (default with a target frame; the clip's own picture) | people (default without one; the people become a depth map and are replaced, the background is kept) | scene (the whole frame becomes a depth map, everything seen comes from the references and prompt).
- `sound`: original (default on nemesis: the clip's own sound is put back on the result) | none (default on nemesis-lite: silent).
- `face_mesh` (nemesis-lite only): boolean, default true with guide people, false with scene and clip. Traces each face in the clip so the new face keeps its expressions, eyes and lips; never shown in the result.
- `resolution` 480p | 720p (default) | 1080p, `prompt` optional. The result keeps the clip's length and shape.
- How long: preparing the clip takes ~15-40 s on nemesis (the target frame) and ~25 s to ~1.5 min on nemesis-lite; the render is most of the wait (an 8 s clip at 480p: ~4-7 min).
- Priced at a Seedance 2.5 edit's rate plus $0.02 a second ($0.08 / $0.17 / $0.36 at 480p / 720p / 1080p), per second of clip in plus video out. On nemesis-lite a clip with nobody in it under guide people is refused (422 invalid_input) and not charged.

## Nemesis Lite input (`"model": "nemesis-lite"`)

- The same input as nemesis, rendered on Wan 3.0: cheaper, a little less faithful on faces and fine detail. `video_url` is a 2-15 s clip.
- Priced at Wan 3.0's rate plus $0.02 a second ($0.06 / $0.09 / $0.16 at 480p / 720p / 1080p), per second of clip in plus video out.

## MCP server (for AI agents)

- Server URL: https://revana.si/api/mcp (Streamable HTTP, POST only, stateless). Not https://revana.si/mcp: that is the setup page.
- Auth: OAuth sign-in (the client gets its own key, revocable in the dashboard), or `Authorization: Bearer sr_live_...` with a dashboard key.
- Claude Code: `claude mcp add --transport http revana https://revana.si/api/mcp`. Codex: `codex mcp add revana --url https://revana.si/api/mcp` then `codex mcp login revana`. claude.ai / Claude Desktop / ChatGPT: add a custom connector with the server URL. Cursor, VS Code, Windsurf, Gemini CLI, n8n: https://revana.si/mcp
- Tools: `list_models` (catalog, options, prices; free), `get_quote` (exact price; free), `generate_video` (any live model, Seedance 2.5 by default; spends credits; needs `idempotency_key`), `get_video` (status; `wait_seconds` up to 40), `get_video_link` (download link, 1 hour), `upload_file` (link for a local file, 14 days), `get_balance`, `quote_nemesis` and `make_nemesis` (restage a clip with your own pictures), `buy_credits` (saved-card packs $10 to $5,000; only if allowed at sign-in and the person asks).
- Flow: get_quote, tell the person the price, generate_video after a yes, get_video with wait_seconds 40 until done, get_video_link. The same idempotency_key never makes or charges twice; failed jobs are free; `max_cost_usd` caps a job at up to $100.
- Limits: 20 calls a second per key in bursts; no daily caps, prepaid credits are the limit.

## Docs

- [Quickstart and reference](https://revana.si/docs.html)
- [OpenAPI 3.1 spec](https://revana.si/openapi.json)
- [Models, options and prices (JSON)](https://revana.si/data/models.json)
- [Pricing](https://revana.si/pricing.html)
- [Seedance 2.5: price per clip, API, free credit and comparisons](https://revana.si/seedance-2-5.md)
  - [Seedance 2.5 image to video](https://revana.si/seedance-2-5/image-to-video.md): animate a picture: first frame, first and last frame, code
  - [Seedance 2.5 reference to video](https://revana.si/seedance-2-5/reference-to-video.md): images, clips and sounds as references, limits and naming
  - [Seedance 2.5 prompt guide](https://revana.si/seedance-2-5/prompt-guide.md): a template and 12 example prompts
  - [Seedance 2.5 cost calculator](https://revana.si/seedance-2-5/price-calculator.md): price per second, per clip, per minute
  - [Can you use Seedance 2.5 for free?](https://revana.si/seedance-2-5/free.md): what is really free, and the cheapest paid clip
  - [Seedance 2.5 in Claude and ChatGPT](https://revana.si/seedance-2-5/claude-mcp.md): the MCP server, set up in a minute
  - [Seedance 2.5 vs Kling 4.0](https://revana.si/seedance-2-5/vs-kling-4-0.md): two 30 second models compared
  - [Seedance 2.5 vs Kling 3.0](https://revana.si/seedance-2-5/vs-kling-3-0.md): length, 4K and price
  - [Seedance 2.5 vs Veo 3.1](https://revana.si/seedance-2-5/vs-veo-3-1.md): long takes against 8 second 4K shots
  - [Seedance 2.5 vs Wan 3.0](https://revana.si/seedance-2-5/vs-wan-3-0.md): references against price
  - [Seedance 2.5 vs MiniMax H3](https://revana.si/seedance-2-5/vs-minimax-h3.md): a free draft model and the 30 second one
- [MCP server for AI agents](https://revana.si/mcp.html): setup for every client, every tool and its parameters, pricing, troubleshooting
- [Tutorial: make your first AI video in the dashboard](https://revana.si/tutorials/first-video): sign in, add credit, pick a model, prompt, Generate; narrated 1-minute video
- [Tutorial: connect Revana to Claude](https://revana.si/tutorials/connect-claude): add the connector, approve, ask for a video; narrated walkthrough
- [Chat completions API for language models (OpenAI-compatible)](https://revana.si/chat)
- [Language models and token prices (JSON)](https://revana.si/data/llms.json)
- [Everything in one file, with every model](https://revana.si/llms-full.txt)

<!-- guides:start (scripts/build-revana-model-pages.mjs) -->
## Guides

- [Midjourney alternatives: 8 image models you can call from one API](https://revana.si/blog/midjourney-alternative): Eight Midjourney alternatives you can call from one API: Nano Banana Pro, GPT Image 2, Seedream, FLUX, Ideogram and more, priced per image, with code.
- [Image editing API: edit photos with a prompt, priced per image](https://revana.si/blog/image-editing-api): Edit photos with a prompt through one API: Nano Banana, GPT Image 2, FLUX Kontext, Qwen Image Edit, Seedream and more. Price per edit and working code.
- [AI avatar API: talking avatars and lip sync from one photo, priced per second](https://revana.si/blog/ai-avatar-api): Turn a photo and a voice track into a talking avatar video with OmniHuman or Kling Avatar, or lip sync a clip you already have. Prices per second and code.
- [Ideogram API pricing: Ideogram V3, V4 and Character per image, with code](https://revana.si/blog/ideogram-api): Ideogram API pricing per image for Ideogram V3, V4 and Character at turbo, balanced and quality, what 1,000 images cost, and code for text in images.
- [Recraft API: Recraft V4.1, V4.1 Utility and Crisp Upscale per image, with code](https://revana.si/blog/recraft-api): Recraft V4.1 API pricing per image at 1K and 2K, V4.1 vs V4.1 Utility, the 14 aspect ratios, Crisp Upscale and working code. No Recraft plan needed.
- [ElevenLabs API pricing: every voice model per 1,000 characters, with code](https://revana.si/blog/elevenlabs-api-pricing): ElevenLabs API pricing in November 2026: V4, V3, Multilingual v2 and Turbo 2.5 per 1,000 characters, what an hour of speech costs, and code. No plan.
- [Kling API pricing: every Kling model per second, with code](https://revana.si/blog/kling-api-pricing): Kling AI API pricing in November 2026: Kling 4.0, 4.0 Flash, 3.0, Omni, Turbo, 2.6 and 2.1 per second, what a 5 and 10 second clip costs, and code.
- [MiniMax API: H3 video, Hailuo and M3 chat, with prices and code](https://revana.si/blog/minimax-api): MiniMax API in November 2026: MiniMax H3 and H3 Max video, Hailuo 2.3 and 02, and M3 chat, with prices, which model to pick and working code.
- [ElevenLabs alternative: the same voices without a plan, or video that speaks](https://revana.si/blog/elevenlabs-alternative): Looking for an ElevenLabs alternative? Use the same voices by API with no subscription, or video models that speak, with prices and code.
- [Midjourney API pricing: relax, fast and turbo per request, with code](https://revana.si/blog/midjourney-api-pricing): Midjourney V7 API pricing in November 2026: relax, fast and turbo prices per request, what 100 and 1,000 requests cost, and working code. No subscription.
- [Cheapest Seedance API in November 2026: Seedance 2.5, 2.0 and 2.0 Fast compared](https://revana.si/blog/cheapest-seedance-api): The cheapest Seedance 2.5, 2.0 and 2.0 Fast API prices per second in November 2026: Revana against Kie.ai, fal, Replicate and ByteDance's own API.
- [Kimi K3 API pricing: K3, K3.1 Pro and Highspeed per token, with code](https://revana.si/blog/kimi-k3-api-pricing): Kimi K3, K3.1 Pro and K3.1 Pro Highspeed API prices per million tokens, 40% below Moonshot's list, with OpenAI SDK code and a price check vs Claude Fable 5.
- [GPT Image 2.6 API: Flare vs Sunburst, price per image and working code](https://revana.si/blog/gpt-image-2-6-api): Call OpenAI's GPT Image 2.6 Flare and Sunburst through one API: price per image at 1K and 2K, how the two versions differ, and working code and MCP setup.
- [AI video generation API cost in 2026: per second, per clip and per minute](https://revana.si/blog/ai-video-generation-api-cost): What AI video generation costs through an API in 2026: price per second, per 10 second clip and per minute of footage for Seedance, Kling, Veo, Wan and MiniMax.
- [Kling 4.0 API: price, 30 second clips and working code](https://revana.si/blog/kling-4-0-api): Call Kuaishou's Kling 4.0 through one API: per-second prices at 720p and 1080p, 30 second clips with sound, reference images, 9:16 and working code.
- [Nano Banana Pro vs Nano Banana 2: which image API to use](https://revana.si/blog/nano-banana-pro-vs-nano-banana-2): Nano Banana Pro or Nano Banana 2? Sizes, aspect ratios, reference images and price per image compared, with which one to use for each job and code.
- [GPT Image 2 vs Nano Banana Pro: price, features and code](https://revana.si/blog/gpt-image-2-vs-nano-banana-pro): GPT Image 2 or Nano Banana Pro? Price per image, sizes, shapes and reference limits compared, plus one script that runs the same prompt on both.
- [Seedream 5.0 API: Pro, Lite and Flash compared, with code](https://revana.si/blog/seedream-5-0-api): Call ByteDance Seedream 5.0 Pro, Lite and Flash, plus Seedream 4.5, through one API: sizes up to 4K, reference edits, price per image and code.
- [Qwen Image API: versions, prices and code](https://revana.si/blog/qwen-image-api): Call Alibaba Qwen Image 3.0, 3.0 Pro, 2.x and Qwen Image Edit through one API: multilingual text, instruction edits, price per image and code.
- [FLUX API: FLUX 3, FLUX 2, Kontext and FLUX 3 Video](https://revana.si/blog/flux-api): Call Black Forest Labs FLUX 3, FLUX 2 Pro and Max, FLUX Kontext and FLUX 3 Video through one API: sizes, reference limits, prices and code.
- [Hailuo API: Hailuo 2.3, Hailuo 02 and MiniMax H3](https://revana.si/blog/hailuo-api): Call MiniMax Hailuo 2.3, Hailuo 02 and MiniMax H3 through one API: lengths, resolutions, per-second prices and working text and image to video code.
- [Vidu API: Vidu Q3 and Q4 models, prices and code](https://revana.si/blog/vidu-api): Call Shengshu Vidu Q3, Q3 Turbo, Q3 Pro, Q3 Mix and Q4 through one API: reference to video, clip lengths, per-second prices and working code.
- [Kling Motion Control API: make a character copy a video](https://revana.si/blog/kling-motion-control-api): Make a character from a picture copy the moves in a video with Kling 3.0 and 2.6 Motion Control: inputs, per-second billing and working code.
- [Suno API: generate songs with Suno V6](https://revana.si/blog/suno-api): Make songs and instrumentals with Suno V6 through a REST API: prompt, style, lyrics and instrumental inputs, two tracks per job, price and code.
- [Gemini Omni API: Google video with sound, prices and code](https://revana.si/blog/gemini-omni-api): Call Google Gemini Omni, Gemini Omni 1.1 Flash and Flash Preview through one API: video with sound up to 4K, per-second prices and working code.
- [How to cut AI video API costs: seven practical ways](https://revana.si/blog/reduce-ai-video-api-costs): Seven ways to spend less on AI video generation APIs: low-resolution drafts, cheaper tiers, silent output, spend caps and free failed jobs, with code.
- [AI video API error codes: what they mean and when to retry](https://revana.si/blog/ai-video-api-error-codes): Every error a video or image generation API call can return, what it means, whether to retry it, and a Python retry loop that never pays twice.
- [AI video API price comparison: Revana against every major API](https://revana.si/blog/ai-video-api-price-comparison): Seedance, Veo 3.1, Kling, Wan, Grok Imagine and Nano Banana API prices on Revana, APIMart, Kie.ai, CometAPI, fal, Replicate, WaveSpeed and more, side by side.
- [A Runware alternative: Revana and Runware prices, model by model](https://revana.si/blog/runware-alternative): The Seedance, Veo 3.1 Fast, Grok Imagine and Nano Banana settings where Revana costs less than Runware, side by side, and how to switch.
- [A PiAPI alternative: Revana and PiAPI prices, model by model](https://revana.si/blog/piapi-alternative): The Seedance, Kling 3.0, Wan and Nano Banana settings where Revana costs less than PiAPI, setting by setting, and how to switch.
- [An AI/ML API alternative: Revana and AI/ML API prices, model by model](https://revana.si/blog/aimlapi-alternative): Seedance, Veo 3.1, Wan, HappyHorse and Hailuo prices on Revana against AI/ML API (aimlapi.com), setting by setting, and how to switch.
- [A WaveSpeed alternative: Revana and WaveSpeed prices, model by model](https://revana.si/blog/wavespeed-ai-alternative): The Seedance, Veo, Grok Imagine and Nano Banana Pro settings where Revana costs less than WaveSpeed AI, side by side, and how to switch.
- [An OpenRouter video API alternative: Revana and OpenRouter prices, model by model](https://revana.si/blog/openrouter-video-api-alternative): The Nano Banana, Grok Imagine, Wan and Veo settings where Revana costs less than OpenRouter's video and image models, and how to switch.
- [A Kie.ai alternative: Revana and Kie.ai prices, model by model](https://revana.si/blog/kie-ai-alternative): The Seedance, Nano Banana and GPT Image 2 settings where Revana costs less than Kie.ai, setting by setting, and how to switch.
- [An APIMart alternative: Revana and APIMart prices, model by model](https://revana.si/blog/apimart-alternative): The Grok Imagine Video, Kling 3.0 and Seedream settings where Revana costs less than APIMart, side by side, and how to switch.
- [A CometAPI alternative: Revana and CometAPI prices, model by model](https://revana.si/blog/cometapi-alternative): The Nano Banana, Grok Imagine and Seedance 2.0 Mini settings where Revana costs less than CometAPI, what only Revana sells, and how to switch.
- [The best AI video generation API in 2026: how to choose](https://revana.si/blog/best-ai-video-generation-api-2026): How to choose an AI video generation API in 2026: Seedance, Veo, Kling, Wan, Hailuo and Grok Imagine compared on length, resolution, sound and price.
- [One API for every AI video model](https://revana.si/blog/one-api-for-ai-video-models): Call Seedance, Veo, Kling, Wan and the major image models with one API key, one request format and one prepaid balance, with automatic route fallback.
- [Seedance 2.0 API tutorial: from first request to finished clip](https://revana.si/blog/seedance-2-0-api): Learn the Seedance 2.0 API step by step: get a key, send a request, poll it and download the clip, with every option, prices and cURL, Python and Node.js code.
- [Cheapest Seedance 2.0, Fast and Mini API: prices per second, compared](https://revana.si/blog/seedance-2-0-api-price-comparison): Seedance 2.0, Fast and Mini price per second at Revana, ByteDance, Runway, Replicate, fal, EvoLink and Atlas Cloud, 480p to 4K, with what a clip costs.
- [A cheaper Replicate alternative for Seedance, Veo and Nano Banana](https://revana.si/blog/replicate-alternative): The Seedance 2.0, Veo 3.1, Nano Banana and Grok Imagine settings where Revana costs less than Replicate, with batch costs and how to switch.
- [A cheaper Runway API alternative for Seedance, Veo, Grok and Wan](https://revana.si/blog/runway-api-alternative): The Seedance 2.5, Seedance 2.0, Veo 3.1 Fast, Grok Imagine and Wan settings where Revana costs less than the Runway API, per second, with batch costs.
- [A cheaper fal.ai alternative for Seedance, Veo and Nano Banana](https://revana.si/blog/fal-ai-alternative): The Seedance, Veo 3.1 and Nano Banana settings where Revana costs less than fal, side by side, what a batch costs, and when fal fits better.
- [Grok Imagine Video API pricing: the cheapest price per second](https://revana.si/blog/grok-imagine-video-api-pricing): Grok Imagine Video and Grok Imagine Video 1.5 price per second at Revana, xAI, fal and Replicate at 480p, 720p and 1080p, with what a clip costs.
- [Wan API pricing: the cheapest Wan 3.0, Wan 2.7 and Wan 2.6 API](https://revana.si/blog/wan-api-pricing): Wan 3.0, Wan 2.7 and Wan 2.6 price per second on Revana against fal and Replicate at 720p and 1080p, with what a clip costs and how to call it.
- [Seedance 2.5 vs Veo 3.1 vs Kling 3.0: API price per second and features](https://revana.si/blog/seedance-2-5-vs-veo-3-1-vs-kling-3-0): Seedance 2.5, Veo 3.1, Veo 3.1 Fast and Kling 3.0 side by side: price per second with sound, clip length, resolution, inputs and what a clip costs.
- [Seedance 2.5 vs Seedance 2.0: what changed, and what each costs per second](https://revana.si/blog/seedance-2-5-vs-seedance-2-0): Seedance 2.5 against Seedance 2.0, 2.0 Fast and 2.0 Mini: price per second at 480p to 4K, clip length, reference inputs and which to use for what.
- [The cheapest AI image generation API in 2026, model by model](https://revana.si/blog/cheapest-ai-image-generation-api): Nano Banana Pro, Nano Banana 2, Seedream, Ideogram and Flux price per image at Revana, Google, fal, Replicate and others, and the cheapest image models.
- [PixVerse V6 API pricing: the cheapest price per second](https://revana.si/blog/pixverse-v6-api-pricing): PixVerse V6 price per second at Revana, fal and Replicate from 360p to 1080p, with and without sound, what a clip costs and how to call it.
- [Cheapest AI video generation API in 2026: price per second, model by model](https://revana.si/blog/cheapest-ai-video-generation-api): The cheapest API for Seedance 2.5, Seedance 2.0, Veo 3.1 and more: price per second at Revana, ByteDance, Google, fal and others, and what a minute costs.
- [Cheapest Nano Banana Pro and Nano Banana 2 API: price per image, compared](https://revana.si/blog/nano-banana-pro-api-price-comparison): Nano Banana Pro and Nano Banana 2 price per image at Revana, Google, EvoLink, fal and Atlas Cloud at 1K, 2K and 4K, and what 1,000 images cost.
- [Seedance 2.5 API prices compared: who is cheapest at each resolution](https://revana.si/blog/seedance-2-5-api-price-comparison): Seedance 2.5 price per second on Revana against Segmind, Runway, EvoLink, WaveSpeed, fal and more, 480p to 1080p, with and without a video input.
- [Seedance 2.5 API tutorial: make 30-second clips with code](https://revana.si/blog/seedance-2-5-api): Learn the Seedance 2.5 API step by step: 30-second clips with sound, 30 reference images, polling and download, with prices and cURL, Python and Node.js code.
- [Veo 3.1 API pricing: Veo 3.1, Fast and Lite compared](https://revana.si/blog/veo-3-1-api-pricing): What Veo 3.1, Veo 3.1 Fast and Veo 3.1 Lite cost per second through one API, with and without sound, plus working cURL, Python and Node.js requests.
- [Kling 3.0 API: pricing, sound and working code](https://revana.si/blog/kling-3-0-api): Call Kuaishou's Kling 3.0 through one API: per-second prices with and without sound, the Omni, Turbo and Motion Control variants, and working code.
- [Wan 3.0 API: 30-second clips, pricing and code](https://revana.si/blog/wan-3-0-api): Call Alibaba's Wan 3.0 and Wan 3.0 Prime through one API: clips from 2 to 30 seconds, five input modes, per-second prices and working code samples.
- [Text to video vs image to video: the five API modes explained](https://revana.si/blog/text-to-video-vs-image-to-video-api): Text to video, image to video, first and last frames, reference images and video to video: what each API mode does, when to use it, and example requests.
- [Nano Banana Pro API: price per image and working code](https://revana.si/blog/nano-banana-pro-api): Call Google's Nano Banana Pro (Gemini 3 Pro Image) through one API: price per image at 1K, 2K and 4K, up to 14 input images, and working code.
- [GPT Image 2 API: pricing, sizes and working code](https://revana.si/blog/gpt-image-2-api): Call OpenAI's GPT Image 2 through one API: price per image at 1K, 2K and 4K, nine aspect ratios, up to 16 input images, and working code samples.

<!-- guides:end -->

<!-- try:start (scripts/build-revana-try-pages.mjs) -->
## Try a model online (no code)

- [Try Seedance 2.5](https://revana.si/try/seedance-2-5): ByteDance, long clips with sound, reference images and video edits; type a prompt in the dashboard, one 4 second clip at 480p $0.36.
- [Try Kling 4.0](https://revana.si/try/kling-4-0): Kuaishou, clips up to 30 seconds with sound and consistent characters; type a prompt in the dashboard, one 3 second clip at 720p $0.41.
- [Try Kling 3.0](https://revana.si/try/kling-3-0): Kuaishou, short cinematic clips, up to 4K; type a prompt in the dashboard, one 3 second clip at 720p $0.27.
- [Try Veo 3.1 Fast](https://revana.si/try/veo-3-1-fast): Google, quick 4 to 8 second clips with sound, up to 4K; type a prompt in the dashboard, one 4 second clip at 720p $0.32.
- [Try Veo 3.1](https://revana.si/try/veo-3-1): Google, realistic physics, dialogue, ambience and music in one pass; type a prompt in the dashboard, one 4 second clip at 720p $0.80.
- [Try Wan 3.0](https://revana.si/try/wan-3-0): Alibaba, low-cost clips up to 30 seconds; type a prompt in the dashboard, one 2 second clip at 480p $0.08.
- [Try Wan 3.0 Prime](https://revana.si/try/wan-3-0-prime): Alibaba, Wan 3.0 Prime clips with sound, 2 to 30 seconds at up to 1080p; type a prompt in the dashboard, one 2 second clip at 480p $0.14.
- [Try MiniMax H3](https://revana.si/try/minimax-h3): MiniMax, clips with sound, up to 2K; type a prompt in the dashboard, one 4 second clip at 768p $0.20.
- [Try MiniMax H3 Max](https://revana.si/try/minimax-h3-max): MiniMax, MiniMax H3 at up to 1080p, from text, a picture or reference images; type a prompt in the dashboard, one 5 second clip at 480p $0.25.
- [Try Gemini Omni 1.2 Flash](https://revana.si/try/gemini-omni-1-2-flash): Google, low-cost clips with sound from 360p to 4K, from text, frames or up to 10 reference images; type a prompt in the dashboard, one 3 second clip at 360p $0.08.
- [Try Gemini Omni 1.1 Flash](https://revana.si/try/gemini-omni-1-1-flash): Google, clips with sound from 360p to 4K, priced per job plus per second; type a prompt in the dashboard, one 4 second clip at 360p $0.18.
- [Try Gemini Omni](https://revana.si/try/gemini-omni): Google, clips with sound up to 4K, from text or a picture; type a prompt in the dashboard, one 4 second clip at 720p $0.40.
- [Try Veo 3.1 Lite](https://revana.si/try/veo-3-1-lite): Google, the lowest-cost Veo 3.1 clips, 4 to 8 seconds with sound; type a prompt in the dashboard, one 4 second clip at 720p $0.20.
- [Try Seedance 2.0](https://revana.si/try/seedance-2-0): ByteDance, multi-shot clips that hold characters and style across cuts, up to 4K; type a prompt in the dashboard, one 4 second clip at 480p $0.32.
- [Try Seedance 2.0 Fast](https://revana.si/try/seedance-2-0-fast): ByteDance, faster, lower-cost Seedance 2.0 clips at 480p or 720p, with sound; type a prompt in the dashboard, one 4 second clip at 480p $0.20.
- [Try Seedance 2.0 Mini](https://revana.si/try/seedance-2-0-mini): ByteDance, the lowest-cost Seedance clips, 4 to 15 seconds at 480p or 720p with sound; type a prompt in the dashboard, one 4 second clip at 480p $0.12.
- [Try Seedance 2.0 Face](https://revana.si/try/seedance-2-0-face): ByteDance, Seedance 2.0 clips with sound from frames or reference images, up to 1080p; type a prompt in the dashboard, one 4 second clip at 480p $0.52.
- [Try Seedance 2.0 Fast Face](https://revana.si/try/seedance-2-0-fast-face): ByteDance, faster Seedance 2.0 Face clips at 480p or 720p, with sound; type a prompt in the dashboard, one 4 second clip at 480p $0.40.
- [Try Wan 2.7](https://revana.si/try/wan-2-7): Alibaba, clips from text, a picture, first and last frames, references or another video, up to 1080p; type a prompt in the dashboard, one 2 second clip at 720p $0.18.
- [Try Wan 2.6 Flash](https://revana.si/try/wan-2-6-flash): Alibaba, animating a picture or restyling a clip at low cost, with optional sound; type a prompt in the dashboard, one 5 second clip at 720p $0.15.
- [Try Wan 2.6 I2V Flash](https://revana.si/try/wan-2-6-i2v-flash): Alibaba, low-cost image to video, 2 to 15 seconds, up to 1080p with optional sound; type a prompt in the dashboard, one 2 second clip at 720p $0.06.
- [Try Kling 2.1 Pro](https://revana.si/try/kling-2-1-pro): Kuaishou, turning a picture into a 5 or 10 second clip; type a prompt in the dashboard, one 5 second clip at standard $0.55.
- [Try Kling 2.1 Standard](https://revana.si/try/kling-2-1-std): Kuaishou, low-cost 720p clips from a picture, 5 or 10 seconds; type a prompt in the dashboard, one 5 second clip at 720p $0.35.
- [Try Kling 3.0 Omni](https://revana.si/try/kling-3-0-omni): Kuaishou, Kling 3.0 from text, frames, references or a video input, up to 4K; type a prompt in the dashboard, one 3 second clip at 720p $0.27.
- [Try Hailuo 02](https://revana.si/try/hailuo-02): MiniMax, low-cost 5 or 10 second clips from 512p to 1080p; type a prompt in the dashboard, one 5 second clip at 512p $0.10.
- [Try PixVerse V6](https://revana.si/try/pixverse-v6): PixVerse, clips from 1 to 15 seconds, 360p to 1080p, with optional sound; type a prompt in the dashboard, one 1 second clip at 360p $0.02.
- [Try Grok Imagine Video](https://revana.si/try/grok-imagine-video): xAI, low-cost clips with sound, 6 to 30 seconds, from text or a picture; type a prompt in the dashboard, one 6 second clip at 480p $0.12.
- [Try Grok Imagine Video 1.5](https://revana.si/try/grok-imagine-video-1-5): xAI, clips with sound from text, frames or reference pictures, 1 to 15 seconds; type a prompt in the dashboard, one 1 second clip at 480p $0.04.
- [Try Flux 3 Video](https://revana.si/try/flux-3-video): Black Forest Labs, clips with sound from text or keyframes, 5 to 20 seconds, with a low-cost draft mode; type a prompt in the dashboard, one 5 second clip at draft $0.30.
- [Try Nano Banana Pro](https://revana.si/try/nano-banana-pro): Google, pictures and edits up to 4K; type a prompt in the dashboard, one picture at 1K $0.04.
- [Try Nano Banana 2](https://revana.si/try/nano-banana-2): Google, fast, low-cost pictures and edits; type a prompt in the dashboard, one picture at 1K $0.03.
- [Try Nano Banana 2.1](https://revana.si/try/nano-banana-2-1): Google, pictures and edits up to 4K, with up to 14 reference images; type a prompt in the dashboard, one picture at 1K $0.03.
- [Try Nano Banana 2 Lite](https://revana.si/try/nano-banana-2-lite): Google, the lowest-cost Nano Banana pictures and edits, at 1K; type a prompt in the dashboard, one picture at 1K $0.02.
- [Try Nano Banana](https://revana.si/try/nano-banana): Google, fast, low-cost pictures and edits that keep a subject consistent; type a prompt in the dashboard, one picture $0.03.
- [Try GPT Image 2](https://revana.si/try/gpt-image-2): OpenAI, pictures with readable text, up to 4K; type a prompt in the dashboard, one picture at 1K $0.03.
- [Try GPT Image 2.6 Flare](https://revana.si/try/gpt-image-2-6-flare): OpenAI, OpenAI pictures and edits at 1K or 2K, with up to 16 reference images; type a prompt in the dashboard, one picture at 1K $0.02.
- [Try GPT Image 2.6 Sunburst](https://revana.si/try/gpt-image-2-6-sunburst): OpenAI, the Sunburst variant of GPT Image 2.6: pictures and edits at 1K or 2K from up to 16 references; type a prompt in the dashboard, one picture at 1K $0.02.
- [Try GPT Image 2.5 Flare](https://revana.si/try/gpt-image-2-5-flare): OpenAI, low-cost OpenAI pictures and edits, with a 4K size and up to 16 reference images; type a prompt in the dashboard, one picture at 1K $0.02.
- [Try GPT Image 2.5 Sunburst](https://revana.si/try/gpt-image-2-5-sunburst): OpenAI, the Sunburst variant of GPT Image 2.5: pictures and edits at 1K, 2K or 4K from up to 16 references; type a prompt in the dashboard, one picture at 1K $0.02.
- [Try Seedream 5.5](https://revana.si/try/seedream-5-5): ByteDance, the newest Seedream: sharp pictures and image edits up to 2K; type a prompt in the dashboard, one picture at 1K $0.02.
- [Try Seedream 5.0 Pro](https://revana.si/try/seedream-5-0-pro): ByteDance, detailed pictures and image edits up to 2K; type a prompt in the dashboard, one picture at 1K $0.04.
- [Try Seedream 5.0 Lite](https://revana.si/try/seedream-5-0-lite): ByteDance, pictures and edits at 2K to 4K; type a prompt in the dashboard, one picture at 2K $0.03.
- [Try Seedream 5.0 Flash](https://revana.si/try/seedream-5-0-flash): ByteDance, low-cost pictures and edits up to 2K; type a prompt in the dashboard, one picture $0.02.
- [Try Seedream 4.5](https://revana.si/try/seedream-4-5): ByteDance, sharp 2K and 4K pictures with strong typography; type a prompt in the dashboard, one picture $0.04.
- [Try Seedream 4.0](https://revana.si/try/seedream-4-0): ByteDance, pictures and edits at 1K to 4K for one flat price; type a prompt in the dashboard, one picture $0.03.
- [Try Seedream 3.0](https://revana.si/try/seedream-3-0): ByteDance, low-cost text to image; type a prompt in the dashboard, one picture $0.03.
- [Try Qwen Image 3.0](https://revana.si/try/qwen-image-3-0): Alibaba, pictures and edits up to 2K; type a prompt in the dashboard, one picture at 1K $0.03.
- [Try Qwen Image 3.0 Pro](https://revana.si/try/qwen-image-3-0-pro): Alibaba, higher-quality Qwen pictures and edits at 1K or 2K; type a prompt in the dashboard, one picture at 1K $0.04.
- [Try Qwen Image Edit](https://revana.si/try/qwen-image-edit): Alibaba, editing a picture you give it; type a prompt in the dashboard, one picture $0.04.
- [Try MiniMax H3 Image](https://revana.si/try/minimax-h3-image): MiniMax, detailed 2K pictures and edits at a low price; type a prompt in the dashboard, one picture at 1K $0.02.
- [Try Wan 2.7 Image Pro](https://revana.si/try/wan-2-7-image-pro): Alibaba, pictures and edits at 1K to 4K, with up to 9 reference images; type a prompt in the dashboard, one picture $0.07.
- [Try Z Image](https://revana.si/try/z-image): Alibaba, very low-cost text to image; type a prompt in the dashboard, one picture $0.01.
- [Try Flux 3 Klein](https://revana.si/try/flux-3-klein): Black Forest Labs, low-cost pictures and edits, up to 4K; type a prompt in the dashboard, one picture at 768SQ $0.01.
- [Try Flux 3](https://revana.si/try/flux-3): Black Forest Labs, pictures and edits from 768 square to 4K, with up to 10 reference images; type a prompt in the dashboard, one picture at 768SQ $0.05.
- [Try Flux 2 Pro](https://revana.si/try/flux-2-pro): Black Forest Labs, photoreal pictures with reliable hands and faces, and multi-reference edits; type a prompt in the dashboard, one picture at 1MP $0.03.
- [Try Flux 2 Max](https://revana.si/try/flux-2-max): Black Forest Labs, the highest-quality Flux 2 pictures and edits, 1 to 4 megapixels; type a prompt in the dashboard, one picture at 1MP $0.07.
- [Try Flux Kontext Pro](https://revana.si/try/flux-kontext-pro): Black Forest Labs, editing a picture with words, or making one from text; type a prompt in the dashboard, one picture $0.04.
- [Try Flux Kontext Max](https://revana.si/try/flux-kontext-max): Black Forest Labs, the higher-quality Flux Kontext edits; type a prompt in the dashboard, one picture $0.07.
- [Try Imagen 4](https://revana.si/try/imagen-4): Google, Google's text to image in five aspect ratios; type a prompt in the dashboard, one picture $0.05.
- [Try Imagen 4 Fast](https://revana.si/try/imagen-4-fast): Google, the lowest-cost Imagen 4 pictures; type a prompt in the dashboard, one picture $0.03.
- [Try Imagen 4 Ultra](https://revana.si/try/imagen-4-ultra): Google, Google's highest-quality Imagen 4 pictures; type a prompt in the dashboard, one picture $0.08.
- [Try Hunyuan Image 3.5 Preview](https://revana.si/try/hunyuan-image-3-5-preview): Tencent, pictures and edits up to 4K, with up to 5 reference images; type a prompt in the dashboard, one picture at 1K $0.03.
- [Try Grok Imagine Image 2.0](https://revana.si/try/grok-imagine-image-2-0): xAI, low-cost pictures and edits up to 2K, in many aspect ratios; type a prompt in the dashboard, one picture at 1K $0.02.
- [Try Midjourney V7](https://revana.si/try/midjourney-v7): Midjourney, Midjourney's distinctive look, from a prompt or a reference picture; type a prompt in the dashboard, one picture at relax $0.06.
- [Try Ideogram V3](https://revana.si/try/ideogram-v3): Ideogram, text and graphic design: posters, logos, packaging and social graphics; type a prompt in the dashboard, one picture at turbo $0.03.
- [Try Ideogram Character](https://revana.si/try/ideogram-character): Ideogram, putting the same character into new pictures from one reference; type a prompt in the dashboard, one picture at turbo $0.08.

<!-- try:end -->
<!-- vs:start (scripts/build-revana-vs-pages.mjs) -->
## Comparisons

- [Revana vs Kie.ai](https://revana.si/vs/kie.md): Revana's list price is below Kie.ai's for 34 settings across 13 models: Seedance 2.5, Wan 3.0, Nano Banana Pro, Nano Banana 2, GPT Image 2, Seedance 2.0, Seedream 5.0 Pro, Seedance 2.0 Fast, Nano Banana 2.1, GPT Image 2.5 Flare, GPT Image 2.5 Sunburst, GPT Image 2.6 Flare and GPT Image 2.6 Sunburst (checked 2026-10-30). Kie.ai lists the same price or less for some settings of 24 models, including Wan 3.0 Prime, Veo 3.1 Fast, Veo 3.1 and Kling 3.0.
- [Revana vs fal](https://revana.si/vs/fal.md): Revana's list price is below fal's for 37 settings across 12 models: Seedance 2.5, Wan 3.0, Nano Banana Pro, Nano Banana 2, Veo 3.1 Fast, Veo 3.1, Seedance 2.0, Wan 2.6, Wan 2.7, Seedance 2.0 Fast, PixVerse V6 and Grok Imagine Video (checked 2026-10-30). fal lists the same price or less for some settings of Kling 3.0, Nano Banana 2, Veo 3.1 Fast and Veo 3.1.
- [Revana vs Replicate](https://revana.si/vs/replicate.md): Revana's list price is below Replicate's for 43 settings across 16 models: Seedance 2.5, Nano Banana Pro, Nano Banana 2, Veo 3.1 Fast, Kling 3.0, Seedance 2.0, Wan 2.6, Seedream 5.0 Pro, Ideogram V3, Seedance 2.0 Fast, Seedance 2.0 Mini, PixVerse V6, Grok Imagine Video, Grok Imagine Video 1.5, Flux 2 Flex and Seedream 5.0 Lite (checked 2026-10-30). Replicate lists the same price or less for some settings of 5 models, including Veo 3.1, Flux Kontext Pro, Veo 3.1 Fast and Seedance 2.0.

<!-- vs:end -->

## Per-model queue

POST https://api.revana.si/v1/queue/revana/{model_id} accepts raw model input and Authorization: Key or Bearer. Returns request_id, status_url and response_url. Poll IN_QUEUE / IN_PROGRESS / COMPLETED; optional fal_webhook delivers an OK/ERROR completion. Use an Idempotency-Key on retries. Same credit holds and account isolation as /v1/generations. Download result URLs with your key. See https://revana.si/queue.md for verified fal aliases, webhook verification, unsupported options and SDK migration details.

<!-- compare:start (scripts/build-revana-model-pages.mjs) -->
## Where Revana has the lowest listed price (checked 2026-10-26)

Lower than every other API we checked, including the model maker's own, at these resolutions:

- GPT Image 1.5 (`gpt-image-1.5`): 1:1-high $0.11 an image, 1:1-medium $0.03 an image, 2:3-high $0.14 an image, 2:3-medium $0.03 an image, 3:2-high $0.14 an image, 3:2-medium $0.03 an image
- Wan 2.6 I2V (`wan-2.6-i2v`): 720p $0.09/s, 1080p $0.14/s
- Grok Imagine Video 1.5 (`grok-imagine-video-1.5`): 480p $0.04/s, 720p $0.07/s, 1080p $0.13/s. Compared: https://revana.si/blog/grok-imagine-video-api-pricing
- Qwen Image 3.0 (`qwen-image-3.0`): 1K $0.03 an image, 2K $0.03 an image
- Seedream 5.0 Lite (`seedream-5.0-lite`): 2K $0.03 an image, 3K $0.03 an image. Compared: https://revana.si/blog/cheapest-ai-image-generation-api
- Nano Banana 2.1 (`nano-banana-2.1`): 4K $0.04 an image
- GPT Image 2.5 Flare (`gpt-image-2.5-flare`): 1K $0.02 an image, 2K $0.02 an image, 4K $0.03 an image
- GPT Image 2.5 Sunburst (`gpt-image-2.5-sunburst`): 1K $0.02 an image, 2K $0.02 an image, 4K $0.03 an image
- GPT Image 2.6 Flare (`gpt-image-2.6-flare`): 1K $0.02 an image, 2K $0.02 an image. Compared: https://revana.si/blog/gpt-image-2-6-api
- GPT Image 2.6 Sunburst (`gpt-image-2.6-sunburst`): 1K $0.02 an image, 2K $0.02 an image
- Grok Imagine Image 2.0 (`grok-imagine-image-2.0`): 1K $0.02 an image, 2K $0.02 an image

## Seedance 2.5 price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p | 1080p |
|---|---|---|---|
| Revana | $0.09 | $0.21 | $0.52 |
| BytePlus ModelArk | $0.1028 | $0.2311 | $0.5686 |
| fal | $0.2205 | $0.473 | $1.164 |
| Runway API | $0.20 | $0.30 | $0.68 |
| EvoLink | $0.138 | $0.296 | $0.739 |
| Atlas Cloud | $0.134 | $0.303 | – |
| Replicate | $0.1028 | $0.2312 | – |
| Segmind | $0.1065 | $0.2389 | $0.5876 |
| WaveSpeed | $0.162 | $0.324 | $0.81 |
| Kie.ai | $0.14 | $0.315 | $0.79 |
| APIMart | $0.0961 | $0.216 | $0.5346 |
| OpenRouter | $0.1028 | $0.2311 | – |
| Together AI | $0.115 | $0.249 | – |
| Vercel AI Gateway | $0.1028 | $0.2311 | $0.5686 |
| Runware | $0.1028 | $0.2311 | $0.6085 |
| PiAPI | $0.15 | $0.35 | $0.80 |
| AI/ML API | $0.13364 | $0.30046 | $0.67603 |
| Higgsfield API | $0.144 | – | $0.7961 |

With a video input (the input clip's seconds are billed too):

| API | 480p | 720p | 1080p |
|---|---|---|---|
| Revana | $0.06 | $0.15 | $0.34 |
| BytePlus ModelArk | $0.0615 | – | – |
| EvoLink | $0.084 | $0.18 | $0.45 |
| WaveSpeed | $0.099 | $0.198 | $0.495 |
| Kie.ai | $0.085 | $0.19 | $0.475 |
| OpenRouter | $0.0615 | – | – |
| Vercel AI Gateway | $0.0615 | – | $0.3402 |
| Runware | $0.0658 | – | $0.364 |
| AI/ML API | $0.07993 | $0.17971 | $0.40435 |

Higgsfield prices are its published maximum-discount API rates; eligibility may require selecting the model, and offers can change. Source: https://open.higgsfield.ai/pricing

Full comparison and sources: https://revana.si/blog/seedance-2-5-api-price-comparison (data: https://revana.si/data/market-prices.json)

## Wan 3.0 price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p | 1080p |
|---|---|---|---|
| Revana | $0.04 | $0.07 | $0.14 |
| fal | $0.05 | $0.10 | $0.20 |
| Runway API | $0.05 | $0.10 | $0.20 |
| CometAPI | – | $0.08 | $0.16 |
| Kie.ai | – | $0.08 | $0.16 |
| WaveSpeed | $0.0475 | $0.095 | $0.19 |
| OpenRouter | $0.05 | $0.10 | $0.20 |
| Vercel AI Gateway | $0.05 | $0.10 | $0.20 |
| Runware | $0.05 | $0.10 | $0.20 |
| AI/ML API | $0.065 | $0.13 | $0.26 |

Full comparison and sources: https://revana.si/blog/wan-api-pricing (data: https://revana.si/data/market-prices.json)

## Wan 3.0 Prime price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p | 1080p |
|---|---|---|---|
| Revana | $0.07 | $0.13 | $0.26 |
| WaveSpeed | $0.07125 | $0.1425 | $0.285 |
| OpenRouter | – | $0.14 | $0.28 |
| Vercel AI Gateway | – | $0.14 | $0.28 |


## Nano Banana Pro price at other APIs (checked 2026-10-26)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K | 4K |
|---|---|---|---|
| Revana | $0.04 | $0.04 | $0.08 |
| Google's own API | $0.134 | $0.134 | $0.24 |
| EvoLink | $0.126 | $0.126 | $0.224 |
| Atlas Cloud | $0.14 | – | – |
| fal | $0.15 | $0.15 | $0.30 |
| Replicate | $0.15 | $0.15 | $0.30 |
| CometAPI | $0.1072 | $0.1072 | $0.192 |
| Kie.ai | $0.09 | $0.09 | $0.12 |
| WaveSpeed | $0.126 | $0.126 | $0.216 |
| OpenRouter | $0.1344 | $0.1344 | $0.24 |
| Together AI | $0.134 | $0.134 | $0.24 |
| Vercel AI Gateway | $0.1344 | $0.1344 | $0.24 |
| Runware | $0.138 | $0.138 | $0.244 |
| PiAPI | $0.105 | $0.105 | $0.18 |

Full comparison and sources: https://revana.si/blog/nano-banana-pro-api-price-comparison (data: https://revana.si/data/market-prices.json)

## Nano Banana 2 price at other APIs (checked 2026-10-26)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K | 4K |
|---|---|---|---|
| Revana | $0.03 | $0.03 | $0.04 |
| Google's own API | $0.067 | $0.101 | $0.151 |
| fal | $0.08 | $0.12 | $0.16 |
| Replicate | $0.067 | $0.101 | $0.151 |
| CometAPI | $0.0536 | $0.0808 | $0.1208 |
| Kie.ai | $0.04 | $0.06 | $0.09 |
| WaveSpeed | $0.063 | $0.0945 | $0.126 |
| OpenRouter | $0.0672 | $0.1008 | $0.1512 |
| Vercel AI Gateway | $0.067 | $0.101 | $0.151 |
| Runware | $0.06895 | $0.10255 | $0.15295 |
| PiAPI | $0.06 | $0.08 | $0.12 |

Full comparison and sources: https://revana.si/blog/nano-banana-pro-api-price-comparison (data: https://revana.si/data/market-prices.json)

## GPT Image 2 price at other APIs (checked 2026-10-28)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K | 4K |
|---|---|---|---|
| Revana | $0.03 | $0.03 | $0.03 |
| Kie.ai | – | $0.05 | $0.08 |
| WaveSpeed | $0.057 | $0.095 | $0.171 |
| PiAPI | $0.032 | – | – |
| OpenAI API | $0.211 | – | – |

OpenAI comparison: 1024 × 1024, high quality, output only; input tokens cost extra. OpenAI offers other sizes and quality levels at different prices. Source: https://developers.openai.com/api/docs/guides/image-generation

Full comparison and sources: https://revana.si/blog/ai-video-api-price-comparison (data: https://revana.si/data/market-prices.json)

## Veo 3.1 Fast price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip; the +audio columns are with sound. Card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 720p+audio | 1080p | 1080p+audio | 4K | 4K+audio |
|---|---|---|---|---|---|---|
| Revana | $0.08 | $0.10 | $0.10 | $0.11 | $0.25 | $0.29 |
| Google's own API | – | – | – | $0.12 | – | $0.30 |
| fal | $0.10 | $0.15 | – | $0.15 | $0.30 | $0.35 |
| Replicate | $0.10 | $0.15 | – | $0.15 | – | – |
| Runway API | $0.10 | $0.15 | – | $0.15 | – | – |
| WaveSpeed | $0.10 | $0.15 | – | $0.15 | – | $0.30 |
| OpenRouter | – | – | – | $0.12 | – | $0.30 |
| Vercel AI Gateway | $0.10 | $0.15 | – | $0.15 | $0.30 | $0.35 |
| Runware | $0.10 | $0.15 | – | $0.15 | $0.30 | $0.35 |
| AI/ML API | $0.13 | $0.195 | $0.13 | $0.195 | – | – |

Full comparison and sources: https://revana.si/blog/veo-3-1-api-pricing (data: https://revana.si/data/market-prices.json)

## Veo 3.1 price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip; the +audio columns are with sound. Card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 720p+audio | 1080p | 1080p+audio | 4K+audio |
|---|---|---|---|---|---|
| Revana | $0.20 | $0.40 | $0.20 | $0.40 | $0.58 |
| Google's own API | – | – | – | – | $0.60 |
| fal | – | – | – | – | $0.60 |
| WaveSpeed | – | – | – | – | $0.60 |
| OpenRouter | – | – | – | – | $0.60 |
| Vercel AI Gateway | – | – | – | – | $0.60 |
| Runware | – | – | – | – | $0.60 |
| AI/ML API | $0.26 | $0.52 | $0.26 | $0.52 | – |

Full comparison and sources: https://revana.si/blog/ai-video-api-price-comparison (data: https://revana.si/data/market-prices.json)

## Kling 3.0 price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip; the +audio columns are with sound. Card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 720p+audio | 1080p | 1080p+audio | 4K | 4K+audio |
|---|---|---|---|---|---|---|
| Revana | $0.09 | $0.13 | $0.12 | $0.17 | $0.42 | $0.42 |
| APIMart | – | – | – | – | $0.4286 | $0.4286 |
| PiAPI | $0.10 | $0.15 | $0.15 | $0.20 | – | – |
| Replicate | – | – | – | $0.336 | – | – |

Full comparison and sources: https://revana.si/blog/ai-video-api-price-comparison (data: https://revana.si/data/market-prices.json)

## Kling 2.6 Pro price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip; the +audio columns are with sound. Card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1080p | 1080p+audio |
|---|---|---|
| Revana | $0.08 | $0.16 |
| AI/ML API | $0.091 | $0.182 |

Full comparison and sources: https://revana.si/blog/aimlapi-alternative (data: https://revana.si/data/market-prices.json)

## Seedance 2.0 price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p | 1080p | 4K |
|---|---|---|---|---|
| Revana | $0.08 | $0.15 | $0.40 | $0.83 |
| BytePlus ModelArk | – | $0.1512 | – | – |
| fal | $0.1345 | $0.3034 | $0.682 | – |
| EvoLink | $0.093 | $0.199 | $0.497 | $1.013 |
| Atlas Cloud | $0.09 | – | – | – |
| Replicate | – | $0.18 | $0.45 | $1.00 |
| Runway API | $0.36 | $0.36 | – | $1.50 |
| Kie.ai | $0.095 | $0.205 | $0.51 | $1.04 |
| WaveSpeed | $0.108 | $0.216 | $0.54 | $1.08 |
| OpenRouter | – | $0.1512 | – | – |
| Vercel AI Gateway | – | $0.1512 | – | – |
| Runware | – | $0.16 | – | $0.8559 |
| PiAPI | $0.10 | $0.20 | $0.50 | – |
| AI/ML API | $0.09243 | – | – | – |
| Higgsfield API | $0.0985 | – | – | $1.0887 |

Higgsfield prices are its published maximum-discount API rates; eligibility may require selecting the model, and offers can change. Source: https://open.higgsfield.ai/pricing

Full comparison and sources: https://revana.si/blog/seedance-2-0-api-price-comparison (data: https://revana.si/data/market-prices.json)

## Wan 2.6 price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 1080p |
|---|---|---|
| Revana | $0.07 | $0.11 |
| fal | $0.10 | $0.15 |
| Replicate | $0.10 | $0.15 |
| WaveSpeed | $0.10 | $0.15 |
| OpenRouter | $0.08 | $0.12 |
| Vercel AI Gateway | $0.10 | $0.15 |
| Runware | $0.10 | $0.15 |
| PiAPI | $0.08 | $0.12 |

Full comparison and sources: https://revana.si/blog/wan-api-pricing (data: https://revana.si/data/market-prices.json)

## Hailuo 2.3 price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 768p | 1080p |
|---|---|---|
| Revana | $0.06 | $0.09 |
| WaveSpeed | – | $0.098 |
| AI/ML API | $0.060667 | $0.106167 |

Full comparison and sources: https://revana.si/blog/ai-video-api-price-comparison (data: https://revana.si/data/market-prices.json)

## GPT Image 1.5 price at other APIs (checked 2026-10-30)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1:1-high | 1:1-medium | 2:3-high | 2:3-medium | 3:2-high | 3:2-medium |
|---|---|---|---|---|---|---|
| Revana | $0.11 | $0.03 | $0.14 | $0.03 | $0.14 | $0.03 |
| OpenAI API | $0.133 | $0.034 | $0.20 | $0.05 | $0.20 | $0.05 |

OpenAI comparison: output-only estimates for the size and quality shown; input tokens cost extra. Source: https://developers.openai.com/api/docs/guides/image-generation

Revana has the lowest listed price per image of every API we checked at 1:1-high ($0.11), 1:1-medium ($0.03), 2:3-high ($0.14), 2:3-medium ($0.03), 3:2-high ($0.14) and 3:2-medium ($0.03).

## Seedream 5.0 Pro price at other APIs (checked 2026-10-26)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 1.5K | 2K |
|---|---|---|---|
| Revana | $0.04 | $0.04 | $0.04 |
| Atlas Cloud | $0.045 | $0.045 | $0.09 |
| Replicate | $0.045 | – | $0.09 |
| Kie.ai | – | – | $0.07 |
| APIMart | – | – | $0.0585 |
| WaveSpeed | $0.0405 | $0.0405 | $0.081 |
| OpenRouter | $0.045 | – | $0.09 |
| Runware | – | $0.04815 | $0.0963 |
| PiAPI | $0.068 | – | $0.136 |

Full comparison and sources: https://revana.si/blog/cheapest-ai-image-generation-api (data: https://revana.si/data/market-prices.json)

## Flux 2 Pro price at other APIs (checked 2026-10-28)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1MP | 2MP | 3MP | 4MP |
|---|---|---|---|---|
| Revana | $0.03 | $0.05 | $0.06 | $0.08 |
| AI/ML API | $0.039 | $0.078 | $0.117 | $0.156 |

Full comparison and sources: https://revana.si/blog/aimlapi-alternative (data: https://revana.si/data/market-prices.json)

## Ideogram V3 price at other APIs (checked 2026-10-28)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | balanced | quality |
|---|---|---|
| Revana | $0.05 | $0.07 |
| Replicate | $0.06 | $0.09 |
| WaveSpeed | $0.06 | $0.09 |
| Runware | $0.06 | $0.09 |

Full comparison and sources: https://revana.si/blog/cheapest-ai-image-generation-api (data: https://revana.si/data/market-prices.json)

## HappyHorse 1.1 price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 1080p |
|---|---|---|
| Revana | $0.15 | $0.19 |
| AI/ML API | $0.182 | $0.234 |

Full comparison and sources: https://revana.si/blog/ai-video-api-price-comparison (data: https://revana.si/data/market-prices.json)

## Wan 2.6 I2V price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 1080p |
|---|---|---|
| Revana | $0.09 | $0.14 |
| OpenRouter | $0.10 | $0.15 |
| Vercel AI Gateway | $0.10 | $0.15 |

Revana has the lowest listed price per second of every API we checked at 720p ($0.09) and 1080p ($0.14).

## Wan 2.7 price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 1080p |
|---|---|---|
| Revana | $0.09 | $0.14 |
| fal | $0.10 | $0.15 |
| OpenRouter | $0.10 | – |
| Vercel AI Gateway | $0.10 | $0.15 |

Full comparison and sources: https://revana.si/blog/wan-api-pricing (data: https://revana.si/data/market-prices.json)

## Seedance 2.0 Fast price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p |
|---|---|---|
| Revana | $0.05 | $0.11 |
| BytePlus ModelArk | $0.0538 | $0.121 |
| fal | $0.1076 | $0.2419 |
| Atlas Cloud | $0.072 | $0.1548 |
| Replicate | $0.07 | $0.15 |
| Runway API | $0.29 | $0.29 |
| Kie.ai | $0.059 | $0.124 |
| WaveSpeed | $0.10 | $0.20 |
| Vercel AI Gateway | $0.0538 | $0.121 |
| Runware | $0.06 | $0.13 |
| AI/ML API | $0.07436 | $0.16003 |

Full comparison and sources: https://revana.si/blog/seedance-2-0-api-price-comparison (data: https://revana.si/data/market-prices.json)

## Seedance 2.0 Mini price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p |
|---|---|---|
| Revana | $0.03 | $0.05 |
| BytePlus ModelArk | $0.0336 | $0.0756 |
| Atlas Cloud | $0.045 | $0.0968 |
| Replicate | $0.04 | $0.09 |
| Runway API | $0.16 | $0.16 |
| CometAPI | $0.032 | $0.064 |
| WaveSpeed | $0.06 | $0.12 |
| OpenRouter | $0.0336 | $0.0756 |
| Vercel AI Gateway | $0.0336 | $0.0756 |
| Runware | $0.036 | $0.081 |
| PiAPI | $0.042 | $0.084 |

Full comparison and sources: https://revana.si/blog/seedance-2-0-api-price-comparison (data: https://revana.si/data/market-prices.json)

## Veo 3.1 Lite price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 1080p |
|---|---|---|
| Revana | $0.05 | $0.06 |
| Google's own API | – | $0.08 |
| fal | – | $0.08 |
| Replicate | – | $0.08 |
| WaveSpeed | – | $0.08 |
| OpenRouter | – | $0.08 |
| Vercel AI Gateway | – | $0.08 |
| Runware | – | $0.08 |
| AI/ML API | $0.065 | $0.104 |

Full comparison and sources: https://revana.si/blog/veo-3-1-api-pricing (data: https://revana.si/data/market-prices.json)

## MiniMax H3 price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 768p | 2K |
|---|---|---|
| Revana | $0.05 | $0.08 |
| Vercel AI Gateway | $0.08 | $0.13 |


## PixVerse V6 price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip; the +audio columns are with sound. Card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 360p | 360p+audio | 540p | 540p+audio | 720p | 720p+audio | 1080p | 1080p+audio |
|---|---|---|---|---|---|---|---|---|
| Revana | $0.02 | $0.03 | $0.03 | $0.04 | $0.04 | $0.05 | $0.08 | $0.10 |
| fal | $0.025 | $0.035 | $0.035 | $0.045 | $0.045 | $0.06 | $0.09 | $0.115 |
| Replicate | $0.05 | $0.07 | $0.07 | $0.09 | $0.09 | $0.12 | $0.18 | $0.23 |
| WaveSpeed | $0.025 | $0.035 | $0.035 | $0.045 | $0.045 | $0.06 | $0.09 | $0.115 |
| Runware | $0.025 | $0.035 | $0.035 | $0.045 | $0.045 | $0.06 | $0.09 | $0.115 |

Full comparison and sources: https://revana.si/blog/pixverse-v6-api-pricing (data: https://revana.si/data/market-prices.json)

## Grok Imagine Video price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p |
|---|---|---|
| Revana | $0.02 | $0.03 |
| xAI's own API | $0.05 | $0.07 |
| fal | $0.05 | $0.07 |
| Replicate | $0.05 | $0.05 |
| CometAPI | $0.04 | $0.056 |
| APIMart | $0.04 | $0.056 |
| WaveSpeed | $0.05 | $0.05 |
| OpenRouter | $0.05 | $0.07 |
| Vercel AI Gateway | $0.05 | $0.07 |
| Runware | $0.05 | $0.07 |

Full comparison and sources: https://revana.si/blog/grok-imagine-video-api-pricing (data: https://revana.si/data/market-prices.json)

## Grok Imagine Video 1.5 price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p | 1080p |
|---|---|---|---|
| Revana | $0.04 | $0.07 | $0.13 |
| xAI's own API | $0.08 | $0.14 | $0.25 |
| fal | $0.08 | $0.14 | $0.25 |
| Replicate | $0.08 | $0.08 | – |
| Runway API | $0.10 | $0.16 | $0.29 |
| CometAPI | $0.064 | $0.112 | $0.20 |
| APIMart | $0.064 | $0.112 | $0.20 |
| WaveSpeed | $0.08 | $0.14 | $0.25 |
| OpenRouter | $0.08 | $0.14 | $0.25 |
| Vercel AI Gateway | $0.08 | $0.14 | $0.25 |
| Runware | $0.08 | $0.14 | $0.25 |

Revana has the lowest listed price per second of every API we checked at 480p ($0.04), 720p ($0.07) and 1080p ($0.13).
Full comparison and sources: https://revana.si/blog/grok-imagine-video-api-pricing (data: https://revana.si/data/market-prices.json)

## Qwen Image 3.0 price at other APIs (checked 2026-10-30)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K |
|---|---|---|
| Revana | $0.03 | $0.03 |
| Higgsfield API | $0.04 | $0.075 |

Higgsfield prices are its published maximum-discount API rates; eligibility may require selecting the model, and offers can change. Source: https://open.higgsfield.ai/pricing

Revana has the lowest listed price per image of every API we checked at 1K ($0.03) and 2K ($0.03).

## Flux 2 Flex price at other APIs (checked 2026-10-28)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1MP | 2MP | 3MP | 4MP |
|---|---|---|---|---|
| Revana | $0.05 | $0.10 | $0.15 | $0.20 |
| Replicate | $0.06 | $0.12 | $0.18 | $0.24 |

Full comparison and sources: https://revana.si/blog/cheapest-ai-image-generation-api (data: https://revana.si/data/market-prices.json)

## Seedream 5.0 Lite price at other APIs (checked 2026-10-30)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 2K | 3K |
|---|---|---|
| Revana | $0.03 | $0.03 |
| Replicate | $0.035 | $0.035 |

Revana has the lowest listed price per image of every API we checked at 2K ($0.03) and 3K ($0.03).
Full comparison and sources: https://revana.si/blog/cheapest-ai-image-generation-api (data: https://revana.si/data/market-prices.json)

## Nano Banana 2.1 price at other APIs (checked 2026-10-30)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 4K |
|---|---|
| Revana | $0.04 |
| Kie.ai | $0.045 |

Revana has the lowest listed price per image of every API we checked at 4K ($0.04).

## GPT Image 2.5 Flare price at other APIs (checked 2026-10-30)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K | 4K |
|---|---|---|---|
| Revana | $0.02 | $0.02 | $0.03 |
| Kie.ai | $0.03 | $0.05 | $0.08 |

Revana has the lowest listed price per image of every API we checked at 1K ($0.02), 2K ($0.02) and 4K ($0.03).

## GPT Image 2.5 Sunburst price at other APIs (checked 2026-10-30)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K | 4K |
|---|---|---|---|
| Revana | $0.02 | $0.02 | $0.03 |
| Kie.ai | $0.03 | $0.05 | $0.08 |

Revana has the lowest listed price per image of every API we checked at 1K ($0.02), 2K ($0.02) and 4K ($0.03).

## GPT Image 2.6 Flare price at other APIs (checked 2026-11-05)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K |
|---|---|---|
| Revana | $0.02 | $0.02 |
| Kie.ai | $0.03 | $0.05 |

Revana has the lowest listed price per image of every API we checked at 1K ($0.02) and 2K ($0.02).
Full comparison and sources: https://revana.si/blog/gpt-image-2-6-api (data: https://revana.si/data/market-prices.json)

## GPT Image 2.6 Sunburst price at other APIs (checked 2026-11-05)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K |
|---|---|---|
| Revana | $0.02 | $0.02 |
| Kie.ai | $0.03 | $0.05 |

Revana has the lowest listed price per image of every API we checked at 1K ($0.02) and 2K ($0.02).

## Grok Imagine Image 2.0 price at other APIs (checked 2026-10-30)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K |
|---|---|---|
| Revana | $0.02 | $0.02 |
| Higgsfield API | $0.04 | $0.08 |

Higgsfield prices are its published maximum-discount API rates; eligibility may require selecting the model, and offers can change. Source: https://open.higgsfield.ai/pricing

Revana has the lowest listed price per image of every API we checked at 1K ($0.02) and 2K ($0.02).
<!-- compare:end -->

<!-- llms:start (scripts/build-revana-llms.mjs) -->
## Language models (chat completions)

POST https://api.revana.si/v1/chat/completions takes the OpenAI chat completions body (streaming included) with your Revana key; any OpenAI SDK works with base URL https://api.revana.si/v1. Billed per token from prepaid credit at the prices below (USD per million tokens, input / output); `usage.cost_usd` on each answer is what it cost. Docs: https://revana.si/chat. Data: https://revana.si/data/llms.json (prices checked 2026-10-30).

- [Claude Fable 5](https://revana.si/llm/claude-fable-5.md) (`claude-fable-5`): Anthropic, $3.50 / $17.50, 1M context
- [Claude Fable 5.1](https://revana.si/llm/claude-fable-5-1.md) (`claude-fable-5-1`): Anthropic, $3.50 / $17.50, 1M context
- [Claude Haiku 4.5](https://revana.si/llm/claude-haiku-4-5-20251001.md) (`claude-haiku-4-5-20251001`): Anthropic, $1.00 / $5.00, 128K context
- [Claude Haiku 4.5 (thinking)](https://revana.si/llm/claude-haiku-4-5-20251001-thinking.md) (`claude-haiku-4-5-20251001-thinking`): Anthropic, $1.00 / $5.00
- [Claude Haiku 5.5](https://revana.si/llm/claude-haiku-5-5.md) (`claude-haiku-5-5`): Anthropic, $0.035 / $0.175, 1M context
- [Claude Opus 4.5](https://revana.si/llm/claude-opus-4-5-20251101.md) (`claude-opus-4-5-20251101`): Anthropic, $5.00 / $25.00
- [Claude Opus 4.5 (thinking)](https://revana.si/llm/claude-opus-4-5-20251101-thinking.md) (`claude-opus-4-5-20251101-thinking`): Anthropic, $5.00 / $25.00
- [Claude Opus 4.6](https://revana.si/llm/claude-opus-4-6.md) (`claude-opus-4-6`): Anthropic, $5.00 / $10.00, 1M context
- [Claude Opus 4.6 (thinking)](https://revana.si/llm/claude-opus-4-6-thinking.md) (`claude-opus-4-6-thinking`): Anthropic, $5.00 / $25.00
- [Claude Opus 4.7](https://revana.si/llm/claude-opus-4-7.md) (`claude-opus-4-7`): Anthropic, $5.00 / $25.00, 128K context
- [Claude Opus 4.8](https://revana.si/llm/claude-opus-4-8.md) (`claude-opus-4-8`): Anthropic, $1.75 / $8.75, 1M context
- [Claude Opus 5](https://revana.si/llm/claude-opus-5.md) (`claude-opus-5`): Anthropic, $1.75 / $8.75, 1M context
- [Claude Opus 5.5](https://revana.si/llm/claude-opus-5-5.md) (`claude-opus-5-5`): Anthropic, $1.40 / $7.00, 1M context
- [Claude Sonnet 4.5](https://revana.si/llm/claude-sonnet-4-5-20250929.md) (`claude-sonnet-4-5-20250929`): Anthropic, $3.00 / $15.00, 128K context
- [Claude Sonnet 4.5 (thinking)](https://revana.si/llm/claude-sonnet-4-5-20250929-thinking.md) (`claude-sonnet-4-5-20250929-thinking`): Anthropic, $3.00 / $15.00
- [Claude Sonnet 4.6](https://revana.si/llm/claude-sonnet-4-6.md) (`claude-sonnet-4-6`): Anthropic, $3.00 / $15.00, 128K context
- [Claude Sonnet 4.6 (thinking)](https://revana.si/llm/claude-sonnet-4-6-thinking.md) (`claude-sonnet-4-6-thinking`): Anthropic, $3.00 / $15.00
- [Claude Sonnet 5](https://revana.si/llm/claude-sonnet-5.md) (`claude-sonnet-5`): Anthropic, $0.70 / $3.50, 1M context
- [Claude Sonnet 5.5](https://revana.si/llm/claude-sonnet-5-5.md) (`claude-sonnet-5-5`): Anthropic, $0.70 / $3.50, 1M context
- [DeepSeek R1](https://revana.si/llm/deepseek-r1.md) (`deepseek-r1`): DeepSeek, $0.56 / $2.24
- [DeepSeek V3.2](https://revana.si/llm/deepseek-v3.2.md) (`deepseek-v3.2`): DeepSeek, $0.257 / $0.3855
- [DeepSeek V4 Flash](https://revana.si/llm/deepseek-v4-flash.md) (`deepseek-v4-flash`): DeepSeek, $0.4286 / $1.2857, 1M context
- [DeepSeek V4 Pro](https://revana.si/llm/deepseek-v4-pro.md) (`deepseek-v4-pro`): DeepSeek, $0.396 / $1.188, 1M context
- [DeepSeek V4.1 Flash](https://revana.si/llm/deepseek-v4.1-flash.md) (`deepseek-v4.1-flash`): DeepSeek, $0.09 / $0.36, 1M context
- [Gemini 3 Flash Preview](https://revana.si/llm/gemini-3-flash-preview.md) (`gemini-3-flash-preview`): Google, $0.50 / $3.00
- [Gemini 3 Flash Preview (no thinking)](https://revana.si/llm/gemini-3-flash-preview-nothinking.md) (`gemini-3-flash-preview-nothinking`): Google, $0.50 / $3.00
- [Gemini 3 Flash Preview (thinking)](https://revana.si/llm/gemini-3-flash-preview-thinking.md) (`gemini-3-flash-preview-thinking`): Google, $0.50 / $3.00
- [Gemini 3 Pro Preview](https://revana.si/llm/gemini-3-pro-preview.md) (`gemini-3-pro-preview`): Google, $2.00 / $12.00
- [Gemini 3 Pro Preview (thinking)](https://revana.si/llm/gemini-3-pro-preview-thinking.md) (`gemini-3-pro-preview-thinking`): Google, $2.00 / $12.00
- [Gemini 3.1 Flash Lite Preview](https://revana.si/llm/gemini-3.1-flash-lite-preview.md) (`gemini-3.1-flash-lite-preview`): Google, $0.25 / $1.50
- [Gemini 3.1 Pro Preview](https://revana.si/llm/gemini-3.1-pro-preview.md) (`gemini-3.1-pro-preview`): Google, $2.00 / $12.00
- [Gemini 3.1 Pro Preview (thinking)](https://revana.si/llm/gemini-3.1-pro-preview-thinking.md) (`gemini-3.1-pro-preview-thinking`): Google, $2.00 / $12.00
- [Gemini 3.5 Flash](https://revana.si/llm/gemini-3.5-flash.md) (`gemini-3.5-flash`): Google, $1.50 / $9.00
- [Gemini 3.5 Flash Lite](https://revana.si/llm/gemini-3.5-flash-lite.md) (`gemini-3.5-flash-lite`): Google, $0.30 / $2.499
- [Gemini 3.6 Flash](https://revana.si/llm/gemini-3.6-flash.md) (`gemini-3.6-flash`): Google, $1.50 / $7.50, 1.05M context
- [Gemini 3.7 Flash](https://revana.si/llm/gemini-3.7-flash.md) (`gemini-3.7-flash`): Google, $0.75 / $3.75, 1.05M context
- [Gemini 3.8 Flash](https://revana.si/llm/gemini-3.8-flash.md) (`gemini-3.8-flash`): Google, $0.225 / $1.125, 1M context
- [Gemini 4 Argon](https://revana.si/llm/gemini-4-argon.md) (`gemini-4-argon`): Google, $0.60 / $3.00, 1M context
- [Kimi K2.5](https://revana.si/llm/kimi-k2.5.md) (`kimi-k2.5`): Kimi, $0.514 / $0.6939
- [Kimi K2.6](https://revana.si/llm/kimi-k2.6.md) (`kimi-k2.6`): Kimi, $0.57 / $2.40, 256K context
- [Kimi K2.7 Code](https://revana.si/llm/kimi-k2.7-code.md) (`kimi-k2.7-code`): Kimi, $0.57 / $2.40, 256K context
- [Kimi K3](https://revana.si/llm/kimi-k3.md) (`kimi-k3`): Kimi, $1.80 / $9.00, 1M context
- [Kimi K3.1 Pro](https://revana.si/llm/kimi-k3.1-pro.md) (`kimi-k3.1-pro`): Kimi, $1.20 / $6.00, 1M context
- [Kimi K3.1 Pro Highspeed](https://revana.si/llm/kimi-k3.1-pro-highspeed.md) (`kimi-k3.1-pro-highspeed`): Kimi, $2.40 / $12.00, 1M context
- [MiniMax M2.5](https://revana.si/llm/minimax-m2.5.md) (`minimax-m2.5`): MiniMax, $0.30 / $1.20
- [MiniMax M3](https://revana.si/llm/minimax-m3.md) (`minimax-m3`): MiniMax, $0.30 / $1.20
- [GPT-4.1](https://revana.si/llm/gpt-4.1.md) (`gpt-4.1`): OpenAI, $2.00 / $8.00, 128K context
- [GPT-4.1 mini](https://revana.si/llm/gpt-4.1-mini.md) (`gpt-4.1-mini`): OpenAI, $0.40 / $1.60, 128K context
- [GPT-4.1 nano](https://revana.si/llm/gpt-4.1-nano.md) (`gpt-4.1-nano`): OpenAI, $0.10 / $0.40, 128K context
- [GPT-4o](https://revana.si/llm/gpt-4o.md) (`gpt-4o`): OpenAI, $2.50 / $10.00, 128K context
- [GPT-4o mini](https://revana.si/llm/gpt-4o-mini.md) (`gpt-4o-mini`): OpenAI, $0.15 / $0.60
- [GPT-5](https://revana.si/llm/gpt-5.md) (`gpt-5`): OpenAI, $1.25 / $10.00, 128K context
- [GPT-5 mini](https://revana.si/llm/gpt-5-mini.md) (`gpt-5-mini`): OpenAI, $0.25 / $2.00, 128K context
- [GPT-5 nano](https://revana.si/llm/gpt-5-nano.md) (`gpt-5-nano`): OpenAI, $0.05 / $0.40, 128K context
- [GPT-5.1](https://revana.si/llm/gpt-5.1.md) (`gpt-5.1`): OpenAI, $1.25 / $10.00, 128K context
- [GPT-5.2](https://revana.si/llm/gpt-5.2.md) (`gpt-5.2`): OpenAI, $1.75 / $14.00, 128K context
- [GPT-5.3 Codex](https://revana.si/llm/gpt-5.3-codex.md) (`gpt-5.3-codex`): OpenAI, $1.75 / $14.00
- [GPT-5.4](https://revana.si/llm/gpt-5.4.md) (`gpt-5.4`): OpenAI, $2.50 / $15.00, 1M context
- [GPT-5.4 mini](https://revana.si/llm/gpt-5.4-mini.md) (`gpt-5.4-mini`): OpenAI, $0.75 / $4.50, 128K context
- [GPT-5.4 nano](https://revana.si/llm/gpt-5.4-nano.md) (`gpt-5.4-nano`): OpenAI, $0.20 / $1.25, 128K context
- [GPT-5.5](https://revana.si/llm/gpt-5.5.md) (`gpt-5.5`): OpenAI, $1.15 / $6.90, 1.05M context
- [GPT-5.6 Luna](https://revana.si/llm/gpt-5.6-luna.md) (`gpt-5.6-luna`): OpenAI, $0.20 / $1.20, 922K context
- [GPT-5.6 Sol](https://revana.si/llm/gpt-5.6-sol.md) (`gpt-5.6-sol`): OpenAI, $0.92 / $4.60, 1.05M context
- [GPT-5.6 Terra](https://revana.si/llm/gpt-5.6-terra.md) (`gpt-5.6-terra`): OpenAI, $0.46 / $2.76, 1.05M context
- [GPT-6 Astra](https://revana.si/llm/gpt-6-astra.md) (`gpt-6-astra`): OpenAI, $2.30 / $11.50, 1.05M context
- [GPT-6 Luna](https://revana.si/llm/gpt-6-luna.md) (`gpt-6-luna`): OpenAI, $0.10 / $0.50, 922K context
- [GPT-6 Sol](https://revana.si/llm/gpt-6-sol.md) (`gpt-6-sol`): OpenAI, $0.46 / $2.30, 1.05M context
- [GPT-6.1 Sol](https://revana.si/llm/gpt-6.1-sol.md) (`gpt-6.1-sol`): OpenAI, $0.46 / $2.30, 1.05M context
- [GPT-6.1 Terra](https://revana.si/llm/gpt-6.1-terra.md) (`gpt-6.1-terra`): OpenAI, $0.23 / $1.15, 1.05M context
- [o3](https://revana.si/llm/o3.md) (`o3`): OpenAI, $2.00 / $8.00
- [o3-mini](https://revana.si/llm/o3-mini.md) (`o3-mini`): OpenAI, $1.10 / $4.40, 128K context
- [o3-mini (high)](https://revana.si/llm/o3-mini-high.md) (`o3-mini-high`): OpenAI, $1.10 / $4.40
- [o3-mini (low)](https://revana.si/llm/o3-mini-low.md) (`o3-mini-low`): OpenAI, $1.10 / $4.40
- [o3-mini (medium)](https://revana.si/llm/o3-mini-medium.md) (`o3-mini-medium`): OpenAI, $1.10 / $4.40
- [o4-mini](https://revana.si/llm/o4-mini.md) (`o4-mini`): OpenAI, $1.10 / $4.40, 128K context
- [Qwen3.6-Flash](https://revana.si/llm/qwen3.6-flash.md) (`qwen3.6-flash`): Qwen, $0.1714 / $1.0284
- [Qwen3.6-Plus](https://revana.si/llm/qwen3.6-plus.md) (`qwen3.6-plus`): Qwen, $0.2858 / $1.7148
- [Qwen3.7-Flash](https://revana.si/llm/qwen3.7-flash.md) (`qwen3.7-flash`): Qwen, $0.0286 / $0.1143, 984K context
- [Qwen3.7-Max](https://revana.si/llm/qwen3.7-max.md) (`qwen3.7-max`): Qwen, $1.7142 / $5.1426
- [Qwen3.7-Plus](https://revana.si/llm/qwen3.7-plus.md) (`qwen3.7-plus`): Qwen, $0.2858 / $1.1432
- [Qwen3.8 2.4T-A95B](https://revana.si/llm/qwen3.8-2.4t-a95b.md) (`qwen3.8-2.4t-a95b`): Qwen, $1.7143 / $5.1429, 984K context
- [Qwen3.8 27B](https://revana.si/llm/qwen3.8-27b.md) (`qwen3.8-27b`): Qwen, $0.4286 / $1.7143, 984K context
- [Qwen3.8-Flash](https://revana.si/llm/qwen3.8-flash.md) (`qwen3.8-flash`): Qwen, $0.1143 / $0.3857, 984K context
- [Qwen3.8-Max](https://revana.si/llm/qwen3.8-max.md) (`qwen3.8-max`): Qwen, $1.20 / $3.60, 1M context
- [Step 3.7 Flash](https://revana.si/llm/step-3.7-flash.md) (`step-3.7-flash`): StepFun, $0.1928 / $1.1568
- [Grok 4.3](https://revana.si/llm/grok-4.3.md) (`grok-4.3`): xAI, $1.25 / $2.50
- [Grok 4.5](https://revana.si/llm/grok-4.5.md) (`grok-4.5`): xAI, $0.60 / $1.80, 500K context
- [Grok 4.6](https://revana.si/llm/grok-4.6.md) (`grok-4.6`): xAI, $2.00 / $6.00, 500K context
- [Grok 4.7](https://revana.si/llm/grok-4.7.md) (`grok-4.7`): xAI, $0.60 / $1.80, 500K context
- [Grok 4.20 (non-reasoning)](https://revana.si/llm/grok-4.20-0309-non-reasoning.md) (`grok-4.20-0309-non-reasoning`): xAI, $1.25 / $2.50
- [Grok 4.20 (reasoning)](https://revana.si/llm/grok-4.20-0309-reasoning.md) (`grok-4.20-0309-reasoning`): xAI, $1.25 / $2.50
- [Grok 4.20 Multi-Agent](https://revana.si/llm/grok-4.20-multi-agent-0309.md) (`grok-4.20-multi-agent-0309`): xAI, $1.25 / $2.50
- [Grok Build 0.1](https://revana.si/llm/grok-build-0.1.md) (`grok-build-0.1`): xAI, $1.00 / $2.00
- [GLM-5](https://revana.si/llm/glm-5.md) (`glm-5`): Zhipu GLM, $1.00 / $1.60
- [GLM-5.1](https://revana.si/llm/glm-5.1.md) (`glm-5.1`): Zhipu GLM, $0.84 / $2.64, 200K context
- [GLM-5.2](https://revana.si/llm/glm-5.2.md) (`glm-5.2`): Zhipu GLM, $0.84 / $2.64, 1M context
- [GLM-5.3](https://revana.si/llm/glm-5.3.md) (`glm-5.3`): Zhipu GLM, $0.84 / $2.64, 1M context
- [GLM-5.3 Flash](https://revana.si/llm/glm-5.3-flash.md) (`glm-5.3-flash`): Zhipu GLM, $0.09 / $0.30, 1M context
- [GLM-5.3 FlashX](https://revana.si/llm/glm-5.3-flashx.md) (`glm-5.3-flashx`): Zhipu GLM, $0.222 / $0.75, 1M context

<!-- llms:end -->

# Every model

## Seedance 2.5 API

> ByteDance's newest video model: up to 30 seconds with sound, reference images, video and audio, and first and last frames. By ByteDance, on Revana from $0.09 per second of video.

- Model id: `seedance-2.5`
- Kind: video
- Maker: ByteDance
- Tasks: Text to video, Image to video, First + last frame, Reference to video, Video to video
- Resolutions: 480p, 720p, 1080p
- Durations: 4 to 30 seconds
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive
- Audio: supported (`"audio": true`)
- Input images: up to 30
- Routes: 6 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/seedance-2.5

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.09 | 9 |
| 720p | $0.21 | 21 |
| 1080p | $0.52 | 52 |
| 480p + with video input | $0.06 | 6 |
| 720p + with video input | $0.15 | 15 |
| 1080p + with video input | $0.34 | 34 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p | 1080p |
|---|---|---|---|
| Revana | $0.09 | $0.21 | $0.52 |
| BytePlus ModelArk | $0.1028 | $0.2311 | $0.5686 |
| fal | $0.2205 | $0.473 | $1.164 |
| Runway API | $0.20 | $0.30 | $0.68 |
| EvoLink | $0.138 | $0.296 | $0.739 |
| Atlas Cloud | $0.134 | $0.303 | – |
| Replicate | $0.1028 | $0.2312 | – |
| Segmind | $0.1065 | $0.2389 | $0.5876 |
| WaveSpeed | $0.162 | $0.324 | $0.81 |
| Kie.ai | $0.14 | $0.315 | $0.79 |
| APIMart | $0.0961 | $0.216 | $0.5346 |
| OpenRouter | $0.1028 | $0.2311 | – |
| Together AI | $0.115 | $0.249 | – |
| Vercel AI Gateway | $0.1028 | $0.2311 | $0.5686 |
| Runware | $0.1028 | $0.2311 | $0.6085 |
| PiAPI | $0.15 | $0.35 | $0.80 |
| AI/ML API | $0.13364 | $0.30046 | $0.67603 |
| Higgsfield API | $0.144 | – | $0.7961 |

With a video input (the input clip's seconds are billed too):

| API | 480p | 720p | 1080p |
|---|---|---|---|
| Revana | $0.06 | $0.15 | $0.34 |
| BytePlus ModelArk | $0.0615 | – | – |
| EvoLink | $0.084 | $0.18 | $0.45 |
| WaveSpeed | $0.099 | $0.198 | $0.495 |
| Kie.ai | $0.085 | $0.19 | $0.475 |
| OpenRouter | $0.0615 | – | – |
| Vercel AI Gateway | $0.0615 | – | $0.3402 |
| Runware | $0.0658 | – | $0.364 |
| AI/ML API | $0.07993 | $0.17971 | $0.40435 |

Higgsfield prices are its published maximum-discount API rates; eligibility may require selecting the model, and offers can change. Source: https://open.higgsfield.ai/pricing

Full comparison and sources: https://revana.si/blog/seedance-2-5-api-price-comparison (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "seedance-2.5",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 4,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 1.3
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.84). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Request fields

| Field | Values | Notes |
|---|---|---|
| `model` | "seedance-2.5" | Required, beside input. |
| `task` | "video_edit" | Only with input.video_url: edit that clip. |
| `input.prompt` | text, up to 30,000 characters | Required: the subject, the action, the camera and the sound. |
| `input.resolution` | 480p, 720p and 1080p | 720p when left out. A second costs 480p $0.09, 720p $0.21, 1080p $0.52. |
| `input.duration` | 4 to 30 seconds | 5 when left out. Not with input.video_url: an edit keeps the clip's length. |
| `input.aspect_ratio` | 16:9, 9:16, 1:1, 4:3, 3:4, 21:9 and adaptive | 16:9 when left out; adaptive takes the shape of your picture. |
| `input.audio` | true or false | false when left out. Sound costs nothing extra. |
| `input.image_url` | an https link | An exact first frame. Not with reference media. |
| `input.end_image_url` | an https link | The last frame; needs input.image_url. |
| `input.image_urls` | up to 30 https links | Reference pictures: characters, products, places. |
| `input.video_url` | an https link | The clip to edit (with task video_edit). |
| `input.video_urls` | up to 10 https links | Reference clips, for motion or camera work. |
| `input.audio_urls` | up to 10 https links | Reference sound: a voice, music or effects. |

### Seedance 2.5 API in Python and Node.js

The request above as a whole script: an `Idempotency-Key` made once, a `max_cost_usd` cap, polling until the video is ready, and `out.mp4` saved. Set `REVANA_API_KEY` first; the Node.js script runs on Node 18 or later as an `.mjs` file.

#### Python

```python
import os, sys, time, uuid, requests

API = "https://api.revana.si/v1"
headers = {"Authorization": f"Bearer {os.environ['REVANA_API_KEY']}"}
body = {
  "model": "seedance-2.5",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 4,
    "aspect_ratio": "16:9",
    "audio": True
  },
  "max_cost_usd": 1.3  # your cap: the most this may cost (it lists at $0.84)
}

## One Idempotency-Key per request, made once. Every resend carries it, so a timeout,
## a 429 or a 5xx can never make (or charge for) a second generation.
key = str(uuid.uuid4())
for attempt in range(1, 6):
    try:
        r = requests.post(f"{API}/generations", json=body, timeout=60,
                          headers={**headers, "Idempotency-Key": key})
        if r.status_code != 429 and r.status_code < 500:
            break
    except requests.RequestException:
        pass  # no answer: it may have been made, so resend with the same key
    if attempt == 5:
        sys.exit(f"Not accepted yet; resend later with Idempotency-Key {key}")
    time.sleep(5 * attempt)
gen = r.json()
if not r.ok:  # refused: nothing was made or charged
    sys.exit(gen["error"]["message"])
gen_id = gen["id"]

## Poll for up to 20 minutes
deadline = time.time() + 20 * 60
while gen["status"] not in ("succeeded", "failed"):
    if time.time() > deadline:
        sys.exit(f"{gen_id} is still {gen['status']}; poll it again later")
    time.sleep(5)
    r = requests.get(f"{API}/generations/{gen_id}", headers=headers, timeout=60)
    if r.status_code == 429 or r.status_code >= 500:
        continue  # a hiccup: poll again
    gen = r.json()
    if not r.ok:
        sys.exit(gen["error"]["message"])

## A failed generation is free. To try it again, make a new key: this one answers with the failure.
if gen["status"] == "failed":
    sys.exit(gen["error"]["message"])

file = requests.get(f"{API}/generations/{gen_id}/content", headers=headers, timeout=300)
open("out.mp4", "wb").write(file.content)
print("saved out.mp4", gen["cost_usd"])
```

#### Node.js

```javascript
import { randomUUID } from "node:crypto";
import { writeFile } from "node:fs/promises";

const API = "https://api.revana.si/v1";
const headers = {
  Authorization: `Bearer ${process.env.REVANA_API_KEY}`,
  "Content-Type": "application/json",
};
const body = JSON.stringify({
  "model": "seedance-2.5",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 4,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 1.3 // your cap: the most this may cost (it lists at $0.84)
});

// One Idempotency-Key per request, made once. Every resend carries it, so a timeout,
// a 429 or a 5xx can never make (or charge for) a second generation.
const key = randomUUID();
let res;
for (let attempt = 1; ; attempt++) {
  res = await fetch(`${API}/generations`, {
    method: "POST",
    headers: { ...headers, "Idempotency-Key": key },
    body,
    signal: AbortSignal.timeout(60000),
  }).catch(() => null); // no answer: it may have been made, so resend with the same key
  if (res && res.status !== 429 && res.status < 500) break;
  if (attempt === 5) throw new Error(`Not accepted yet; resend later with Idempotency-Key ${key}`);
  await new Promise((r) => setTimeout(r, attempt * 5000));
}
let gen = await res.json();
if (!res.ok) throw new Error(gen.error.message); // refused: nothing was made or charged
const id = gen.id;

// Poll for up to 20 minutes
const deadline = Date.now() + 20 * 60 * 1000;
while (!["succeeded", "failed"].includes(gen.status)) {
  if (Date.now() > deadline) throw new Error(`${id} is still ${gen.status}; poll it again later`);
  await new Promise((r) => setTimeout(r, 5000));
  res = await fetch(`${API}/generations/${id}`, { headers });
  if (res.status === 429 || res.status >= 500) continue; // a hiccup: poll again
  gen = await res.json();
  if (!res.ok) throw new Error(gen.error.message);
}
// A failed generation is free. To try it again, make a new key: this one answers with the failure.
if (gen.status === "failed") throw new Error(gen.error.message);

const file = await fetch(`${API}/generations/${id}/content`, { headers });
await writeFile("out.mp4", Buffer.from(await file.arrayBuffer()));
console.log("saved out.mp4", gen.cost_usd);
```

### Request bodies

#### Text to video

```json
{
  "model": "seedance-2.5",
  "input": {
    "prompt": "A barista pours latte art in a sunlit cafe, close-up, then a wide shot of the counter; milk steaming and quiet chatter",
    "resolution": "720p",
    "duration": 5,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 1.6
}
```

#### Image to video (exact first frame)

```json
{
  "model": "seedance-2.5",
  "input": {
    "prompt": "The woman in the photo turns to the camera and smiles; slow push-in",
    "resolution": "720p",
    "duration": 5,
    "aspect_ratio": "16:9",
    "audio": true,
    "image_url": "https://example.com/first-frame.jpg"
  },
  "max_cost_usd": 1.6
}
```

#### First and last frame

```json
{
  "model": "seedance-2.5",
  "input": {
    "prompt": "The sketch fills with colour until it becomes the finished painting",
    "resolution": "720p",
    "duration": 5,
    "aspect_ratio": "16:9",
    "audio": true,
    "image_url": "https://example.com/sketch.jpg",
    "end_image_url": "https://example.com/painting.jpg"
  },
  "max_cost_usd": 1.6
}
```

#### Reference pictures

```json
{
  "model": "seedance-2.5",
  "input": {
    "prompt": "The man from the first picture opens the bottle from the second picture on the rooftop from the third",
    "resolution": "720p",
    "duration": 10,
    "aspect_ratio": "16:9",
    "audio": true,
    "image_urls": [
      "https://example.com/man.jpg",
      "https://example.com/bottle.jpg",
      "https://example.com/rooftop.jpg"
    ]
  },
  "max_cost_usd": 3.2
}
```

#### Edit a video

```json
{
  "model": "seedance-2.5",
  "task": "video_edit",
  "input": {
    "prompt": "Make it night, with neon signs reflecting on the wet street",
    "video_url": "https://example.com/clip.mp4",
    "resolution": "720p"
  },
  "max_cost_usd": 4.5
}
```

### Seedance 2.5, Seedance 2.0, Seedance 2.0 Fast and Seedance 2.0 Mini compared

| Model | Model id | From, a second | Up to | Longest | Reference pictures | Edits a video |
|---|---|---|---|---|---|---|
| Seedance 2.5 | `seedance-2.5` | $0.09 | 1080p | 30 s | 30 | yes |
| Seedance 2.0 | `seedance-2.0` | $0.08 | 4K | 15 s | 9 | no |
| Seedance 2.0 Fast | `seedance-2.0-fast` | $0.05 | 720p | 15 s | 9 | no |
| Seedance 2.0 Mini | `seedance-2.0-mini` | $0.03 | 720p | 15 s | 9 | no |

### Questions

**Is Seedance 2.5 available through an API?** Yes. Send "model": "seedance-2.5" to POST https://api.revana.si/v1/generations with your Revana API key. Revana has 6 routes for it: a job starts on the cheapest and moves to the next if one fails, and you are charged once, only for the clip you get.

**How much does a 30 second Seedance 2.5 video cost?** $2.70 at 480p, $6.30 at 720p and $15.60 at 1080p, sound included. With a video input (a clip to edit or reference clips) the rate a second is $0.06 at 480p, $0.15 at 720p and $0.34 at 1080p, and the input clip's seconds are billed too.

**Can Seedance 2.5 edit a video?** Yes. Send "task": "video_edit" with input.video_url (the clip) and a prompt saying what to change. The result keeps the clip's length, so leave input.duration out. It is billed at the video-input rate for the seconds in and out: a 10 second edit at 720p costs $3.00.

**What is the difference between Seedance 2.5 and Seedance 2.0?** On Revana, Seedance 2.5 makes clips of up to 30 seconds (Seedance 2.0: 15), takes up to 30 reference pictures (Seedance 2.0: 9) and goes up to 1080p (Seedance 2.0: 4K). Only Seedance 2.5 edits a video and takes reference video and audio clips here. A second costs $0.09, $0.21 and $0.52 on Seedance 2.5 at 480p, 720p and 1080p, against $0.08, $0.15 and $0.40 on Seedance 2.0; sound costs nothing extra on either. Pick Seedance 2.5 for long clips, many references or a video edit, Seedance 2.0 for 4K.

**Can I try Seedance 2.5 for free?** Yes, your first video. Start from https://revana.si/try/seedance-2-5, sign up and save a card (saving it charges nothing): we credit the cost of one 4 second Seedance 2.5 clip at 480p ($0.36) to your new account. One per account, card and email address.


### Similar models

- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image
- [ByteDance Lip Sync](https://revana.si/models/bytedance-lipsync.md): from $0.05 per second of video

## Seedance 2.5 Fast API

> Seedance 2.5 on dedicated, faster compute: your renders start straight away on capacity kept for this model instead of queueing behind shared traffic, so each video comes back sooner and large batches run side by side. 480p, 720p and 1080p, up to 30 seconds with sound, reference images and audio, and first and last frames. By ByteDance, on Revana from $0.10 per second of video.

- Model id: `seedance-2.5-fast`
- Kind: video
- Maker: ByteDance
- Tasks: Text to video, Image to video, First + last frame, Reference to video, Video to video
- Resolutions: 480p, 720p, 1080p
- Durations: 4 to 30 seconds
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive
- Audio: supported (`"audio": true`)
- Input images: up to 30
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/seedance-2.5-fast

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.10 | 10 |
| 720p | $0.245 | 25 |
| 1080p | $0.55 | 55 |
| 480p + with video input | $0.06 | 6 |
| 720p + with video input | $0.15 | 15 |
| 1080p + with video input | $0.34 | 34 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "seedance-2.5-fast",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 4,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 1.5
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.98). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image
- [ByteDance Lip Sync](https://revana.si/models/bytedance-lipsync.md): from $0.05 per second of video

## Wan 3.0 API

> Alibaba's Wan 3.0: text to video, image to video, first and last frames, reference to video and video to video, up to 1080p, with sound. By Alibaba, on Revana from $0.04 per second of video.

- Model id: `wan-3.0`
- Kind: video
- Maker: Alibaba
- Tasks: Text to video, Image to video, First + last frame, Reference to video, Video to video
- Resolutions: 480p, 720p, 1080p
- Durations: 2 to 30 seconds
- Aspect ratios: adaptive, 16:9, 4:3, 1:1, 3:4, 9:16, 21:9
- Audio: supported (`"audio": true`)
- Input images: up to 10
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/wan-3.0

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.04 | 4 |
| 720p | $0.07 | 7 |
| 1080p | $0.14 | 14 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p | 1080p |
|---|---|---|---|
| Revana | $0.04 | $0.07 | $0.14 |
| fal | $0.05 | $0.10 | $0.20 |
| Runway API | $0.05 | $0.10 | $0.20 |
| CometAPI | – | $0.08 | $0.16 |
| Kie.ai | – | $0.08 | $0.16 |
| WaveSpeed | $0.0475 | $0.095 | $0.19 |
| OpenRouter | $0.05 | $0.10 | $0.20 |
| Vercel AI Gateway | $0.05 | $0.10 | $0.20 |
| Runware | $0.05 | $0.10 | $0.20 |
| AI/ML API | $0.065 | $0.13 | $0.26 |

Full comparison and sources: https://revana.si/blog/wan-api-pricing (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "wan-3.0",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 2,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.21
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.14). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video
- [Wan 2.2 Animate](https://revana.si/models/wan-2.2-animate.md): from $0.08 per second of video

## Wan 3.0 Prime API

> Alibaba's Wan 3.0 Prime: text to video, image to video, first and last frames and reference to video, up to 1080p, with sound. By Alibaba, on Revana from $0.07 per second of video.

- Model id: `wan-3.0-prime`
- Kind: video
- Maker: Alibaba
- Tasks: Text to video, Image to video, First + last frame, Reference to video
- Resolutions: 480p, 720p, 1080p
- Durations: 2 to 30 seconds
- Aspect ratios: adaptive, 16:9, 4:3, 1:1, 3:4, 9:16
- Audio: supported (`"audio": true`)
- Input images: up to 10
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/wan-3.0-prime

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.07 | 7 |
| 720p | $0.13 | 13 |
| 1080p | $0.26 | 26 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p | 1080p |
|---|---|---|---|
| Revana | $0.07 | $0.13 | $0.26 |
| WaveSpeed | $0.07125 | $0.1425 | $0.285 |
| OpenRouter | – | $0.14 | $0.28 |
| Vercel AI Gateway | – | $0.14 | $0.28 |


### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "wan-3.0-prime",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 2,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.39
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.26). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video
- [Wan 2.2 Animate](https://revana.si/models/wan-2.2-animate.md): from $0.08 per second of video

## Nemesis (aka Genjutsu) API

> Nemesis, aka Genjutsu: send a clip and up to 8 pictures, get the same moves, timing and camera played by your people, products and places. Billed on the clip in plus the video out. By Revana, on Revana from $0.08 per second of video.

- Model id: `nemesis`
- Kind: video
- Maker: Revana
- Tasks: Video to video
- Resolutions: 480p, 720p, 1080p
- Durations: 4 to 30 seconds
- Audio: silent only
- Input images: up to 8
- Routes: 6 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/nemesis

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.08 | 8 |
| 720p | $0.17 | 17 |
| 1080p | $0.36 | 36 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "nemesis",
  "input": {
    "video_url": "https://example.com/dance.mp4",
    "image_urls": [
      "https://example.com/me.jpg"
    ],
    "sound": "original",
    "resolution": "720p"
  },
  "max_cost_usd": 5.1
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.17 a second in and out: $3.40 for a 10 s clip). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nemesis Lite](https://revana.si/models/nemesis-lite.md): from $0.06 per second of video
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video

## Nemesis Lite API

> Nemesis Lite: Nemesis on Wan 3.0, for less. Send a clip of up to 15 seconds and up to 8 pictures, get the same moves, timing and camera played by your people, products and places. Billed on the clip in plus the video out. By Revana, on Revana from $0.06 per second of video.

- Model id: `nemesis-lite`
- Kind: video
- Maker: Revana
- Tasks: Video to video
- Resolutions: 480p, 720p, 1080p
- Durations: 2 to 15 seconds
- Audio: silent only
- Input images: up to 8
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/nemesis-lite

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.06 | 6 |
| 720p | $0.09 | 9 |
| 1080p | $0.16 | 16 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "nemesis-lite",
  "input": {
    "video_url": "https://example.com/dance.mp4",
    "image_urls": [
      "https://example.com/me.jpg"
    ],
    "sound": "original",
    "resolution": "720p"
  },
  "max_cost_usd": 2.7
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.09 a second in and out: $1.80 for a 10 s clip). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nemesis (aka Genjutsu)](https://revana.si/models/nemesis.md): from $0.08 per second of video
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video

## Nano Banana Pro API

> Google's top image model. Accurate text in images, multi-image composition and precise edits, up to 4K. By Google, on Revana from $0.04 per image.

- Model id: `nano-banana-pro`
- Kind: image
- Maker: Google
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K, 4K
- Aspect ratios: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 5:4, 4:5, 21:9
- Input images: up to 14
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/nano-banana-pro

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.04 | 4 |
| 2K | $0.04 | 4 |
| 4K | $0.08 | 8 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-26)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K | 4K |
|---|---|---|---|
| Revana | $0.04 | $0.04 | $0.08 |
| Google's own API | $0.134 | $0.134 | $0.24 |
| EvoLink | $0.126 | $0.126 | $0.224 |
| Atlas Cloud | $0.14 | – | – |
| fal | $0.15 | $0.15 | $0.30 |
| Replicate | $0.15 | $0.15 | $0.30 |
| CometAPI | $0.1072 | $0.1072 | $0.192 |
| Kie.ai | $0.09 | $0.09 | $0.12 |
| WaveSpeed | $0.126 | $0.126 | $0.216 |
| OpenRouter | $0.1344 | $0.1344 | $0.24 |
| Together AI | $0.134 | $0.134 | $0.24 |
| Vercel AI Gateway | $0.1344 | $0.1344 | $0.24 |
| Runware | $0.138 | $0.138 | $0.244 |
| PiAPI | $0.105 | $0.105 | $0.18 |

Full comparison and sources: https://revana.si/blog/nano-banana-pro-api-price-comparison (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "nano-banana-pro",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.06
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.04). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video
- [Veo 3.1 Fast](https://revana.si/models/veo-3.1-fast.md): from $0.08 per second of video
- [Veo 3.1](https://revana.si/models/veo-3.1.md): from $0.20 per second of video
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image
- [Gemini Omni](https://revana.si/models/gemini-omni.md): from $0.10 per second of video

## Nano Banana 2 API

> Google's Nano Banana 2: text to image and image editing, up to 4K. By Google, on Revana from $0.03 per image.

- Model id: `nano-banana-2`
- Kind: image
- Maker: Google
- Tasks: Text to image, Image to image
- Resolutions: 0.5K, 1K, 2K, 4K
- Aspect ratios: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 5:4, 4:5, 21:9, 1:4, 4:1, 1:8, 8:1
- Input images: up to 14
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/nano-banana-2

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 0.5K | $0.07 | 7 |
| 1K | $0.03 | 3 |
| 2K | $0.03 | 3 |
| 4K | $0.04 | 4 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-26)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K | 4K |
|---|---|---|---|
| Revana | $0.03 | $0.03 | $0.04 |
| Google's own API | $0.067 | $0.101 | $0.151 |
| fal | $0.08 | $0.12 | $0.16 |
| Replicate | $0.067 | $0.101 | $0.151 |
| CometAPI | $0.0536 | $0.0808 | $0.1208 |
| Kie.ai | $0.04 | $0.06 | $0.09 |
| WaveSpeed | $0.063 | $0.0945 | $0.126 |
| OpenRouter | $0.0672 | $0.1008 | $0.1512 |
| Vercel AI Gateway | $0.067 | $0.101 | $0.151 |
| Runware | $0.06895 | $0.10255 | $0.15295 |
| PiAPI | $0.06 | $0.08 | $0.12 |

Full comparison and sources: https://revana.si/blog/nano-banana-pro-api-price-comparison (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "nano-banana-2",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video
- [Veo 3.1 Fast](https://revana.si/models/veo-3.1-fast.md): from $0.08 per second of video
- [Veo 3.1](https://revana.si/models/veo-3.1.md): from $0.20 per second of video
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image
- [Gemini Omni](https://revana.si/models/gemini-omni.md): from $0.10 per second of video

## GPT Image 2 API

> OpenAI's GPT Image 2: text to image and image editing, up to 4K. By OpenAI, on Revana from $0.03 per image.

- Model id: `gpt-image-2`
- Kind: image
- Maker: OpenAI
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K, 4K
- Aspect ratios: 1:1, 3:4, 4:3, 2:3, 3:2, 9:16, 16:9, 21:9, 5:2
- Input images: up to 16
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/gpt-image-2

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.03 | 3 |
| 2K | $0.03 | 3 |
| 4K | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-28)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K | 4K |
|---|---|---|---|
| Revana | $0.03 | $0.03 | $0.03 |
| Kie.ai | – | $0.05 | $0.08 |
| WaveSpeed | $0.057 | $0.095 | $0.171 |
| PiAPI | $0.032 | – | – |
| OpenAI API | $0.211 | – | – |

OpenAI comparison: 1024 × 1024, high quality, output only; input tokens cost extra. OpenAI offers other sizes and quality levels at different prices. Source: https://developers.openai.com/api/docs/guides/image-generation

Full comparison and sources: https://revana.si/blog/ai-video-api-price-comparison (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "gpt-image-2",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [GPT Image 1.5](https://revana.si/models/gpt-image-1.5.md): from $0.01 per image
- [GPT 4o Image](https://revana.si/models/gpt-4o-image.md): from $0.04 per image
- [GPT Image 1](https://revana.si/models/gpt-image-1.md): from $0.01 per image
- [GPT Image 2.5 Flare](https://revana.si/models/gpt-image-2.5-flare.md): from $0.02 per image
- [GPT Image 2.5 Sunburst](https://revana.si/models/gpt-image-2.5-sunburst.md): from $0.02 per image
- [GPT Image 2.6 Flare](https://revana.si/models/gpt-image-2.6-flare.md): from $0.02 per image

## Soul 2 API

> Higgsfield's Soul 2: fashion-grade, true-to-life photos of people from a prompt, at 720p or 1080p, in seven shapes from 9:16 to 16:9. By Higgsfield, on Revana from $0.01 per image.

- Model id: `soul-2`
- Kind: image
- Maker: Higgsfield
- Tasks: Text to image
- Resolutions: 720p, 1080p
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/soul-2

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 720p | $0.01 | 1 |
| 1080p | $0.01 | 1 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "soul-2",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "720p"
  },
  "max_cost_usd": 0.02
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.01). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Genjutsu Motion Transfer](https://revana.si/models/genjutsu-motion-transfer.md): from $0.21 per second of video
- [Genjutsu Object Swap](https://revana.si/models/genjutsu-object-swap.md): from $0.21 per second of video
- [Marketing Studio Image](https://revana.si/models/marketing-studio-image.md): from $0.04 per image
- [Soul](https://revana.si/models/soul.md): from $0.13 per image
- [Soul Cinema](https://revana.si/models/soul-cinema.md): from $0.01 per image
- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image

## Gemini Omni 1.2 Flash API

> Google's newest Omni video model: 3 to 15 seconds with sound, from text, a first frame, a first and last frame, or up to 10 reference images, at 360p to 4K. By Google, on Revana from $0.024 per second of video.

- Model id: `gemini-omni-1.2-flash`
- Kind: video
- Maker: Google
- Tasks: Text to video, Image to video, First + last frame, Reference to video
- Resolutions: 360p, 720p, 1080p, 4K
- Durations: 3 to 15 seconds
- Aspect ratios: 16:9, 9:16
- Audio: supported (`"audio": true`)
- Input images: up to 10
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/gemini-omni-1.2-flash

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 360p | $0.024 | 3 |
| 720p | $0.072 | 8 |
| 1080p | $0.109 | 11 |
| 4K | $0.217 | 22 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "gemini-omni-1.2-flash",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 3,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.33
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.22). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [Veo 3.1 Fast](https://revana.si/models/veo-3.1-fast.md): from $0.08 per second of video
- [Veo 3.1](https://revana.si/models/veo-3.1.md): from $0.20 per second of video
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image
- [Gemini Omni](https://revana.si/models/gemini-omni.md): from $0.10 per second of video

## Veo 3.1 Fast API

> The quicker, cheaper tier of Veo 3.1. Cinematic motion with synchronized sound, built for iterating on shots before you commit to a final render. By Google, on Revana from $0.08 per second of video.

- Model id: `veo-3.1-fast`
- Kind: video
- Maker: Google
- Tasks: Text to video, Image to video, First + last frame, Reference to video
- Resolutions: 720p, 1080p, 4K
- Durations: 4 s, 6 s, 8 s
- Aspect ratios: 16:9, 9:16, Auto
- Audio: supported (`"audio": true`)
- Input images: up to 3
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/veo-3.1-fast

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.08 | 8 |
| 720p + with audio | $0.10 | 10 |
| 1080p | $0.10 | 10 |
| 1080p + with audio | $0.11 | 11 |
| 4K | $0.25 | 25 |
| 4K + with audio | $0.29 | 29 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip; the +audio columns are with sound. Card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 720p+audio | 1080p | 1080p+audio | 4K | 4K+audio |
|---|---|---|---|---|---|---|
| Revana | $0.08 | $0.10 | $0.10 | $0.11 | $0.25 | $0.29 |
| Google's own API | – | – | – | $0.12 | – | $0.30 |
| fal | $0.10 | $0.15 | – | $0.15 | $0.30 | $0.35 |
| Replicate | $0.10 | $0.15 | – | $0.15 | – | – |
| Runway API | $0.10 | $0.15 | – | $0.15 | – | – |
| WaveSpeed | $0.10 | $0.15 | – | $0.15 | – | $0.30 |
| OpenRouter | – | – | – | $0.12 | – | $0.30 |
| Vercel AI Gateway | $0.10 | $0.15 | – | $0.15 | $0.30 | $0.35 |
| Runware | $0.10 | $0.15 | – | $0.15 | $0.30 | $0.35 |
| AI/ML API | $0.13 | $0.195 | $0.13 | $0.195 | – | – |

Full comparison and sources: https://revana.si/blog/veo-3-1-api-pricing (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "veo-3.1-fast",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 8,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 1.2
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.80). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video
- [Veo 3.1](https://revana.si/models/veo-3.1.md): from $0.20 per second of video
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image
- [Gemini Omni](https://revana.si/models/gemini-omni.md): from $0.10 per second of video

## Veo 3.1 API

> Google's flagship video model. Strong prompt adherence, realistic physics and native dialogue, ambience and music in the same pass. By Google, on Revana from $0.20 per second of video.

- Model id: `veo-3.1`
- Kind: video
- Maker: Google
- Tasks: Text to video, Image to video, First + last frame
- Resolutions: 720p, 1080p, 4K
- Durations: 4 s, 6 s, 8 s
- Aspect ratios: 16:9, 9:16
- Audio: supported (`"audio": true`)
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/veo-3.1

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.20 | 20 |
| 720p + with audio | $0.40 | 40 |
| 1080p | $0.20 | 20 |
| 1080p + with audio | $0.40 | 40 |
| 4K | $0.40 | 40 |
| 4K + with audio | $0.58 | 58 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip; the +audio columns are with sound. Card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 720p+audio | 1080p | 1080p+audio | 4K+audio |
|---|---|---|---|---|---|
| Revana | $0.20 | $0.40 | $0.20 | $0.40 | $0.58 |
| Google's own API | – | – | – | – | $0.60 |
| fal | – | – | – | – | $0.60 |
| WaveSpeed | – | – | – | – | $0.60 |
| OpenRouter | – | – | – | – | $0.60 |
| Vercel AI Gateway | – | – | – | – | $0.60 |
| Runware | – | – | – | – | $0.60 |
| AI/ML API | $0.26 | $0.52 | $0.26 | $0.52 | – |

Full comparison and sources: https://revana.si/blog/ai-video-api-price-comparison (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "veo-3.1",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 8,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 4.8
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $3.20). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video
- [Veo 3.1 Fast](https://revana.si/models/veo-3.1-fast.md): from $0.08 per second of video
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image
- [Gemini Omni](https://revana.si/models/gemini-omni.md): from $0.10 per second of video

## Kling 4.0 API

> Kuaishou's newest Kling: up to 30 seconds at 720p or 1080p, always with stereo sound, from text, a first and last frame, or up to 10 reference images. By Kuaishou, on Revana from $0.135 per second of video.

- Model id: `kling-4.0`
- Kind: video
- Maker: Kuaishou
- Tasks: Text to video, Image to video, First + last frame, Reference to video, Video to video
- Resolutions: 720p, 1080p
- Durations: 3 to 30 seconds
- Aspect ratios: 16:9, 9:16, 1:1, 21:9
- Audio: supported (`"audio": true`)
- Input images: up to 10
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/kling-4.0

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.135 | 14 |
| 720p + with video input | $0.08 | 8 |
| 1080p | $0.20 | 20 |
| 1080p + with video input | $0.12 | 12 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-4.0",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 3,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.62
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.41). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video
- [Kling 2.1 Standard](https://revana.si/models/kling-2.1-std.md): from $0.07 per second of video

## Kling 4.0 Flash API

> The fast, low-cost Kling 4.0: 3 to 20 seconds at 720p, always with sound, from text, a first frame, or up to 10 reference images. By Kuaishou, on Revana $0.10 per second of video.

- Model id: `kling-4.0-flash`
- Kind: video
- Maker: Kuaishou
- Tasks: Text to video, Image to video, Reference to video
- Resolutions: 720p
- Durations: 3 to 20 seconds
- Aspect ratios: 16:9, 9:16, 1:1, 21:9
- Audio: supported (`"audio": true`)
- Input images: up to 10
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/kling-4.0-flash

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.10 | 10 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-4.0-flash",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 3,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.45
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.30). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video
- [Kling 2.1 Standard](https://revana.si/models/kling-2.1-std.md): from $0.07 per second of video

## Kling 3.0 API

> Kuaishou's Kling 3.0: text to video, image to video and first and last frames, up to 4K, with sound. By Kuaishou, on Revana from $0.09 per second of video.

- Model id: `kling-3.0`
- Kind: video
- Maker: Kuaishou
- Tasks: Text to video, Image to video, First + last frame
- Resolutions: 720p, 1080p, 4K
- Durations: 3 to 15 seconds
- Audio: supported (`"audio": true`)
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/kling-3.0

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.09 | 9 |
| 720p + with audio | $0.13 | 13 |
| 1080p | $0.12 | 12 |
| 1080p + with audio | $0.17 | 17 |
| 4K | $0.42 | 42 |
| 4K + with audio | $0.42 | 42 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip; the +audio columns are with sound. Card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 720p+audio | 1080p | 1080p+audio | 4K | 4K+audio |
|---|---|---|---|---|---|---|
| Revana | $0.09 | $0.13 | $0.12 | $0.17 | $0.42 | $0.42 |
| APIMart | – | – | – | – | $0.4286 | $0.4286 |
| PiAPI | $0.10 | $0.15 | $0.15 | $0.20 | – | – |
| Replicate | – | – | – | $0.336 | – | – |

Full comparison and sources: https://revana.si/blog/ai-video-api-price-comparison (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-3.0",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 3,
    "audio": true
  },
  "max_cost_usd": 0.59
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.39). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video
- [Kling 2.1 Standard](https://revana.si/models/kling-2.1-std.md): from $0.07 per second of video

## Kling 2.6 Pro API

> Kling's pro tier: crisp 1080p, expressive faces and fluid camera moves. Sound is optional and priced separately. By Kuaishou, on Revana from $0.08 per second of video.

- Model id: `kling-2.6-pro`
- Kind: video
- Maker: Kuaishou
- Tasks: Text to video, Image to video, First + last frame
- Resolutions: 1080p
- Durations: 5 s, 10 s
- Audio: supported (`"audio": true`)
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/kling-2.6-pro

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 1080p | $0.08 | 8 |
| 1080p + with audio | $0.16 | 16 |
| 1080p + with audio + with voice | $0.19 | 19 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip; the +audio columns are with sound. Card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1080p | 1080p+audio |
|---|---|---|
| Revana | $0.08 | $0.16 |
| AI/ML API | $0.091 | $0.182 |

Full comparison and sources: https://revana.si/blog/aimlapi-alternative (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-2.6-pro",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "1080p",
    "duration": 5,
    "audio": true
  },
  "max_cost_usd": 1.2
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.80). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video
- [Kling 2.1 Standard](https://revana.si/models/kling-2.1-std.md): from $0.07 per second of video

## Seedance 2.0 API

> ByteDance's multi-shot video model. Holds characters and style across cuts, follows first and last frames, and accepts reference images. By ByteDance, on Revana from $0.08 per second of video.

- Model id: `seedance-2.0`
- Kind: video
- Maker: ByteDance
- Tasks: Text to video, Image to video, First + last frame, Reference to video
- Resolutions: 480p, 720p, 1080p, 4K
- Durations: 4 to 15 seconds
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive
- Audio: supported (`"audio": true`)
- Input images: up to 9
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/seedance-2.0

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.08 | 8 |
| 720p | $0.15 | 15 |
| 1080p | $0.40 | 40 |
| 4K | $0.83 | 83 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p | 1080p | 4K |
|---|---|---|---|---|
| Revana | $0.08 | $0.15 | $0.40 | $0.83 |
| BytePlus ModelArk | – | $0.1512 | – | – |
| fal | $0.1345 | $0.3034 | $0.682 | – |
| EvoLink | $0.093 | $0.199 | $0.497 | $1.013 |
| Atlas Cloud | $0.09 | – | – | – |
| Replicate | – | $0.18 | $0.45 | $1.00 |
| Runway API | $0.36 | $0.36 | – | $1.50 |
| Kie.ai | $0.095 | $0.205 | $0.51 | $1.04 |
| WaveSpeed | $0.108 | $0.216 | $0.54 | $1.08 |
| OpenRouter | – | $0.1512 | – | – |
| Vercel AI Gateway | – | $0.1512 | – | – |
| Runware | – | $0.16 | – | $0.8559 |
| PiAPI | $0.10 | $0.20 | $0.50 | – |
| AI/ML API | $0.09243 | – | – | – |
| Higgsfield API | $0.0985 | – | – | $1.0887 |

Higgsfield prices are its published maximum-discount API rates; eligibility may require selecting the model, and offers can change. Source: https://open.higgsfield.ai/pricing

Full comparison and sources: https://revana.si/blog/seedance-2-0-api-price-comparison (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "seedance-2.0",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 4,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.9
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.60). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Request fields

| Field | Values | Notes |
|---|---|---|
| `model` | "seedance-2.0" | Required, beside input. |
| `input.prompt` | text | The subject, the action, the camera and the sound. |
| `input.resolution` | 480p, 720p, 1080p and 4K | A second costs 480p $0.08, 720p $0.15, 1080p $0.40, 4K $0.83. |
| `input.duration` | 4 to 15 seconds | 5 when left out. You pay for the seconds you ask for. |
| `input.aspect_ratio` | 16:9, 9:16, 1:1, 4:3, 3:4, 21:9 and adaptive | adaptive takes the shape of your picture. |
| `input.audio` | true or false | Sound made with the picture; it costs nothing extra. |
| `input.image_url` | an https link | The first frame (image to video). |
| `input.end_image_url` | an https link | The last frame; needs input.image_url. |
| `input.image_urls` | up to 9 https links | Reference pictures; send these or input.image_url, not both. |

### Seedance 2.0 API in Python and Node.js

The request above as a whole script: an `Idempotency-Key` made once, a `max_cost_usd` cap, polling until the video is ready, and `out.mp4` saved. Set `REVANA_API_KEY` first; the Node.js script runs on Node 18 or later as an `.mjs` file.

#### Python

```python
import os, sys, time, uuid, requests

API = "https://api.revana.si/v1"
headers = {"Authorization": f"Bearer {os.environ['REVANA_API_KEY']}"}
body = {
  "model": "seedance-2.0",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 4,
    "aspect_ratio": "16:9",
    "audio": True
  },
  "max_cost_usd": 0.9  # your cap: the most this may cost (it lists at $0.60)
}

## One Idempotency-Key per request, made once. Every resend carries it, so a timeout,
## a 429 or a 5xx can never make (or charge for) a second generation.
key = str(uuid.uuid4())
for attempt in range(1, 6):
    try:
        r = requests.post(f"{API}/generations", json=body, timeout=60,
                          headers={**headers, "Idempotency-Key": key})
        if r.status_code != 429 and r.status_code < 500:
            break
    except requests.RequestException:
        pass  # no answer: it may have been made, so resend with the same key
    if attempt == 5:
        sys.exit(f"Not accepted yet; resend later with Idempotency-Key {key}")
    time.sleep(5 * attempt)
gen = r.json()
if not r.ok:  # refused: nothing was made or charged
    sys.exit(gen["error"]["message"])
gen_id = gen["id"]

## Poll for up to 20 minutes
deadline = time.time() + 20 * 60
while gen["status"] not in ("succeeded", "failed"):
    if time.time() > deadline:
        sys.exit(f"{gen_id} is still {gen['status']}; poll it again later")
    time.sleep(5)
    r = requests.get(f"{API}/generations/{gen_id}", headers=headers, timeout=60)
    if r.status_code == 429 or r.status_code >= 500:
        continue  # a hiccup: poll again
    gen = r.json()
    if not r.ok:
        sys.exit(gen["error"]["message"])

## A failed generation is free. To try it again, make a new key: this one answers with the failure.
if gen["status"] == "failed":
    sys.exit(gen["error"]["message"])

file = requests.get(f"{API}/generations/{gen_id}/content", headers=headers, timeout=300)
open("out.mp4", "wb").write(file.content)
print("saved out.mp4", gen["cost_usd"])
```

#### Node.js

```javascript
import { randomUUID } from "node:crypto";
import { writeFile } from "node:fs/promises";

const API = "https://api.revana.si/v1";
const headers = {
  Authorization: `Bearer ${process.env.REVANA_API_KEY}`,
  "Content-Type": "application/json",
};
const body = JSON.stringify({
  "model": "seedance-2.0",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 4,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.9 // your cap: the most this may cost (it lists at $0.60)
});

// One Idempotency-Key per request, made once. Every resend carries it, so a timeout,
// a 429 or a 5xx can never make (or charge for) a second generation.
const key = randomUUID();
let res;
for (let attempt = 1; ; attempt++) {
  res = await fetch(`${API}/generations`, {
    method: "POST",
    headers: { ...headers, "Idempotency-Key": key },
    body,
    signal: AbortSignal.timeout(60000),
  }).catch(() => null); // no answer: it may have been made, so resend with the same key
  if (res && res.status !== 429 && res.status < 500) break;
  if (attempt === 5) throw new Error(`Not accepted yet; resend later with Idempotency-Key ${key}`);
  await new Promise((r) => setTimeout(r, attempt * 5000));
}
let gen = await res.json();
if (!res.ok) throw new Error(gen.error.message); // refused: nothing was made or charged
const id = gen.id;

// Poll for up to 20 minutes
const deadline = Date.now() + 20 * 60 * 1000;
while (!["succeeded", "failed"].includes(gen.status)) {
  if (Date.now() > deadline) throw new Error(`${id} is still ${gen.status}; poll it again later`);
  await new Promise((r) => setTimeout(r, 5000));
  res = await fetch(`${API}/generations/${id}`, { headers });
  if (res.status === 429 || res.status >= 500) continue; // a hiccup: poll again
  gen = await res.json();
  if (!res.ok) throw new Error(gen.error.message);
}
// A failed generation is free. To try it again, make a new key: this one answers with the failure.
if (gen.status === "failed") throw new Error(gen.error.message);

const file = await fetch(`${API}/generations/${id}/content`, { headers });
await writeFile("out.mp4", Buffer.from(await file.arrayBuffer()));
console.log("saved out.mp4", gen.cost_usd);
```

### Request bodies

#### Text to video

```json
{
  "model": "seedance-2.0",
  "input": {
    "prompt": "A barista pours latte art in a sunlit cafe, close-up, then a wide shot of the counter; milk steaming and quiet chatter",
    "resolution": "720p",
    "duration": 5,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 1.2
}
```

#### Image to video (exact first frame)

```json
{
  "model": "seedance-2.0",
  "input": {
    "prompt": "The woman in the photo turns to the camera and smiles; slow push-in",
    "resolution": "720p",
    "duration": 5,
    "aspect_ratio": "16:9",
    "audio": true,
    "image_url": "https://example.com/first-frame.jpg"
  },
  "max_cost_usd": 1.2
}
```

#### First and last frame

```json
{
  "model": "seedance-2.0",
  "input": {
    "prompt": "The sketch fills with colour until it becomes the finished painting",
    "resolution": "720p",
    "duration": 5,
    "aspect_ratio": "16:9",
    "audio": true,
    "image_url": "https://example.com/sketch.jpg",
    "end_image_url": "https://example.com/painting.jpg"
  },
  "max_cost_usd": 1.2
}
```

#### Reference pictures

```json
{
  "model": "seedance-2.0",
  "input": {
    "prompt": "The man from the first picture opens the bottle from the second picture on the rooftop from the third",
    "resolution": "720p",
    "duration": 10,
    "aspect_ratio": "16:9",
    "audio": true,
    "image_urls": [
      "https://example.com/man.jpg",
      "https://example.com/bottle.jpg",
      "https://example.com/rooftop.jpg"
    ]
  },
  "max_cost_usd": 2.3
}
```

### Seedance 2.0, Seedance 2.0 Fast, Seedance 2.0 Mini and Seedance 2.5 compared

| Model | Model id | From, a second | Up to | Longest | Reference pictures | Edits a video |
|---|---|---|---|---|---|---|
| Seedance 2.0 | `seedance-2.0` | $0.08 | 4K | 15 s | 9 | no |
| Seedance 2.0 Fast | `seedance-2.0-fast` | $0.05 | 720p | 15 s | 9 | no |
| Seedance 2.0 Mini | `seedance-2.0-mini` | $0.03 | 720p | 15 s | 9 | no |
| Seedance 2.5 | `seedance-2.5` | $0.09 | 1080p | 30 s | 30 | yes |

### Questions

**Does Seedance 2.0 have an API?** Yes. On Revana it is the model id seedance-2.0 on POST https://api.revana.si/v1/generations, the same request shape as every other model. Sign in with Google or an email link, add prepaid credit in US dollars and make an API key: there is no waitlist, application or monthly fee.

**Is the Seedance 2.0 API free?** Your first video can be. Start from https://revana.si/try/seedance-2-0, sign up and save a card (saving it charges nothing): we credit the cost of one 4 second Seedance 2.0 clip at 480p ($0.32) to your new account. One per account, card and email address. After that it is pay as you go from $0.08 a second, and a job that fails costs nothing.

**Can I use the Seedance 2.0 API outside China?** Yes. Revana is a US company: you sign in with Google or an email link and pay by card in US dollars, with no Chinese phone number, ID or business licence.

**What is the difference between Seedance 2.0 and Seedance 2.5?** On Revana, Seedance 2.0 makes clips of up to 15 seconds (Seedance 2.5: 30), takes up to 9 reference pictures (Seedance 2.5: 30) and goes up to 4K (Seedance 2.5: 1080p). Only Seedance 2.5 edits a video and takes reference video and audio clips here. A second costs $0.08, $0.15 and $0.40 on Seedance 2.0 at 480p, 720p and 1080p, against $0.09, $0.21 and $0.52 on Seedance 2.5; sound costs nothing extra on either. Pick Seedance 2.0 for 4K, Seedance 2.5 for long clips, many references or a video edit.

**What is the cheapest way to use Seedance 2.0?** Seedance 2.0 Mini, from $0.03 a second, then Seedance 2.0 Fast from $0.05 a second, then Seedance 2.0 from $0.08 a second. All of them take the same inputs (text, a first frame, first and last frames, up to 9 reference pictures); Seedance 2.0 Mini and Seedance 2.0 Fast stop at 720p. A 4 second clip at 480p costs $0.12 on Seedance 2.0 Mini, $0.20 on Seedance 2.0 Fast and $0.32 on Seedance 2.0.

**How long can a Seedance 2.0 video be?** 4 to 15 seconds in one clip, and you pay for the seconds you ask for: 15 seconds costs $1.20 at 480p, $2.25 at 720p, $6.00 at 1080p and $12.45 at 4K. For longer clips, Seedance 2.5 goes up to 30 seconds.


### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image
- [ByteDance Lip Sync](https://revana.si/models/bytedance-lipsync.md): from $0.05 per second of video

## Wan 2.6 API

> Alibaba's video family. Solid motion at a low price, with optional audio and multi-shot prompts. By Alibaba, on Revana from $0.07 per second of video.

- Model id: `wan-2.6`
- Kind: video
- Maker: Alibaba
- Tasks: Text to video, Image to video, Video to video
- Resolutions: 720p, 1080p
- Durations: 5 s, 10 s, 15 s
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4
- Audio: supported (`"audio": true`)
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/wan-2.6

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.07 | 7 |
| 1080p | $0.11 | 11 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 1080p |
|---|---|---|
| Revana | $0.07 | $0.11 |
| fal | $0.10 | $0.15 |
| Replicate | $0.10 | $0.15 |
| WaveSpeed | $0.10 | $0.15 |
| OpenRouter | $0.08 | $0.12 |
| Vercel AI Gateway | $0.10 | $0.15 |
| Runware | $0.10 | $0.15 |
| PiAPI | $0.08 | $0.12 |

Full comparison and sources: https://revana.si/blog/wan-api-pricing (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "wan-2.6",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 5,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.53
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.35). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video
- [Wan 2.2 Animate](https://revana.si/models/wan-2.2-animate.md): from $0.08 per second of video

## Hailuo 2.3 API

> MiniMax's video model, known for dynamic action, dance and physical motion. By MiniMax, on Revana from $0.06 per second of video.

- Model id: `hailuo-2.3`
- Kind: video
- Maker: MiniMax
- Tasks: Text to video, Image to video
- Resolutions: 768p, 1080p
- Durations: 6 s, 10 s
- Audio: silent only
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/hailuo-2.3

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 768p | $0.06 | 6 |
| 1080p | $0.09 | 9 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 768p | 1080p |
|---|---|---|
| Revana | $0.06 | $0.09 |
| WaveSpeed | – | $0.098 |
| AI/ML API | $0.060667 | $0.106167 |

Full comparison and sources: https://revana.si/blog/ai-video-api-price-comparison (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "hailuo-2.3",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "1080p",
    "duration": 6
  },
  "max_cost_usd": 0.81
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.54). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Hailuo 02](https://revana.si/models/hailuo-02.md): from $0.02 per second of video
- [Hailuo 02 Pro](https://revana.si/models/hailuo-02-pro.md): from $0.06 per second of video
- [Hailuo 02 Standard](https://revana.si/models/hailuo-02-std.md): from $0.02 per second of video
- [Hailuo 2.3 Fast](https://revana.si/models/hailuo-2.3-fast.md): from $0.04 per second of video
- [Hailuo 2.3 Standard](https://revana.si/models/hailuo-2.3-std.md): from $0.04 per second of video
- [MiniMax H3](https://revana.si/models/minimax-h3.md): from $0.05 per second of video

## Nano Banana API

> Fast, low-cost image generation and conversational editing. Keeps a subject consistent across a series of edits. By Google, on Revana $0.03 per image.

- Model id: `nano-banana`
- Kind: image
- Maker: Google
- Tasks: Text to image, Image to image
- Input images: up to 14
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/nano-banana

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "nano-banana",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video
- [Veo 3.1 Fast](https://revana.si/models/veo-3.1-fast.md): from $0.08 per second of video
- [Veo 3.1](https://revana.si/models/veo-3.1.md): from $0.20 per second of video
- [Gemini Omni](https://revana.si/models/gemini-omni.md): from $0.10 per second of video

## GPT Image 1.5 API

> OpenAI's image model. Follows long, detailed prompts closely and renders legible text, layouts and edits. By OpenAI, on Revana from $0.01 per image. The lowest listed GPT Image 1.5 price we found at 1:1-high, 1:1-medium, 2:3-high, 2:3-medium, 3:2-high and 3:2-medium (checked 2026-10-26).

- Model id: `gpt-image-1.5`
- Kind: image
- Maker: OpenAI
- Tasks: Text to image, Image to image
- Aspect ratios: 1:1, 3:2, 2:3
- Input images: up to 16
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/gpt-image-1.5

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1:1-high | $0.11 | 11 |
| 1:1-low | $0.01 | 1 |
| 1:1-medium | $0.03 | 3 |
| 2:3-high | $0.14 | 14 |
| 2:3-low | $0.02 | 2 |
| 2:3-medium | $0.03 | 3 |
| 3:2-high | $0.14 | 14 |
| 3:2-low | $0.02 | 2 |
| 3:2-medium | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-30)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1:1-high | 1:1-medium | 2:3-high | 2:3-medium | 3:2-high | 3:2-medium |
|---|---|---|---|---|---|---|
| Revana | $0.11 | $0.03 | $0.14 | $0.03 | $0.14 | $0.03 |
| OpenAI API | $0.133 | $0.034 | $0.20 | $0.05 | $0.20 | $0.05 |

OpenAI comparison: output-only estimates for the size and quality shown; input tokens cost extra. Source: https://developers.openai.com/api/docs/guides/image-generation

Revana has the lowest listed price per image of every API we checked at 1:1-high ($0.11), 1:1-medium ($0.03), 2:3-high ($0.14), 2:3-medium ($0.03), 3:2-high ($0.14) and 3:2-medium ($0.03).

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "gpt-image-1.5",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "aspect_ratio": "1:1"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [GPT 4o Image](https://revana.si/models/gpt-4o-image.md): from $0.04 per image
- [GPT Image 1](https://revana.si/models/gpt-image-1.md): from $0.01 per image
- [GPT Image 2.5 Flare](https://revana.si/models/gpt-image-2.5-flare.md): from $0.02 per image
- [GPT Image 2.5 Sunburst](https://revana.si/models/gpt-image-2.5-sunburst.md): from $0.02 per image
- [GPT Image 2.6 Flare](https://revana.si/models/gpt-image-2.6-flare.md): from $0.02 per image

## Seedream 5.5 API

> ByteDance's newest Seedream image model: 1K to 2K pictures from a prompt, or edits from up to 10 reference images, with sharp text and consistent subjects. By ByteDance, on Revana from $0.02 per image.

- Model id: `seedream-5.5`
- Kind: image
- Maker: ByteDance
- Tasks: Text to image, Image to image
- Resolutions: 1K, 1.5K, 2K
- Aspect ratios: 1:1, 4:3, 3:4, 16:9, 9:16, 3:2, 2:3, 21:9, 2:1, 1:2, 9:21
- Input images: up to 10
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/seedream-5.5

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.02 | 2 |
| 1.5K | $0.02 | 2 |
| 2K | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "seedream-5.5",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.03
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.02). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image
- [ByteDance Lip Sync](https://revana.si/models/bytedance-lipsync.md): from $0.05 per second of video

## Seedream 4.5 API

> ByteDance's image model. Sharp 2K and 4K output, strong typography and reference-guided edits. By ByteDance, on Revana $0.04 per image.

- Model id: `seedream-4.5`
- Kind: image
- Maker: ByteDance
- Tasks: Text to image, Image to image
- Resolutions: 2K, 4K
- Input images: up to 14
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/seedream-4.5

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.04 | 4 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "seedream-4.5",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "2K"
  },
  "max_cost_usd": 0.06
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.04). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image
- [ByteDance Lip Sync](https://revana.si/models/bytedance-lipsync.md): from $0.05 per second of video

## Seedream 5.0 Pro API

> ByteDance's Seedream 5.0 Pro: text to image and image editing, up to 2K. By ByteDance, on Revana from $0.04 per image.

- Model id: `seedream-5.0-pro`
- Kind: image
- Maker: ByteDance
- Tasks: Text to image, Image to image
- Resolutions: 1K, 1.5K, 2K
- Aspect ratios: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 5:4, 4:5, 21:9
- Input images: up to 10
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/seedream-5.0-pro

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.04 | 4 |
| 1.5K | $0.04 | 4 |
| 2K | $0.04 | 4 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-26)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 1.5K | 2K |
|---|---|---|---|
| Revana | $0.04 | $0.04 | $0.04 |
| Atlas Cloud | $0.045 | $0.045 | $0.09 |
| Replicate | $0.045 | – | $0.09 |
| Kie.ai | – | – | $0.07 |
| APIMart | – | – | $0.0585 |
| WaveSpeed | $0.0405 | $0.0405 | $0.081 |
| OpenRouter | $0.045 | – | $0.09 |
| Runware | – | $0.04815 | $0.0963 |
| PiAPI | $0.068 | – | $0.136 |

Full comparison and sources: https://revana.si/blog/cheapest-ai-image-generation-api (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "seedream-5.0-pro",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.06
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.04). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [ByteDance Lip Sync](https://revana.si/models/bytedance-lipsync.md): from $0.05 per second of video

## Flux 2 Pro API

> Black Forest Labs' production model. Photoreal detail, reliable hands and faces, and multi-reference editing. By Black Forest Labs, on Revana from $0.03 per image.

- Model id: `flux-2-pro`
- Kind: image
- Maker: Black Forest Labs
- Tasks: Text to image, Image to image
- Resolutions: 1MP, 2MP, 3MP, 4MP
- Aspect ratios: auto, 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 21:9, 9:21
- Input images: up to 8
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/flux-2-pro

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1MP | $0.03 | 3 |
| 2MP | $0.05 | 5 |
| 3MP | $0.06 | 6 |
| 4MP | $0.08 | 8 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-28)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1MP | 2MP | 3MP | 4MP |
|---|---|---|---|---|
| Revana | $0.03 | $0.05 | $0.06 | $0.08 |
| AI/ML API | $0.039 | $0.078 | $0.117 | $0.156 |

Full comparison and sources: https://revana.si/blog/aimlapi-alternative (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "flux-2-pro",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1MP",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Flux 3 Video](https://revana.si/models/flux-3-video.md): from $0.06 per second of video
- [Flux 2 Flex](https://revana.si/models/flux-2-flex.md): from $0.05 per image
- [Flux 2 Max](https://revana.si/models/flux-2-max.md): from $0.07 per image
- [Flux 3](https://revana.si/models/flux-3.md): from $0.05 per image
- [Flux 3 Klein](https://revana.si/models/flux-3-klein.md): from $0.01 per image
- [Flux Kontext Max](https://revana.si/models/flux-kontext-max.md): from $0.07 per image

## Midjourney V7 API

> Midjourney's distinctive aesthetic, from a prompt or a reference image. By Midjourney, on Revana from $0.06 per image.

- Model id: `midjourney-v7`
- Kind: image
- Maker: Midjourney
- Tasks: Text to image, Image to image
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/midjourney-v7

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| fast | $0.07 | 7 |
| relax | $0.06 | 6 |
| turbo | $0.13 | 13 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "midjourney-v7",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light"
  },
  "max_cost_usd": 0.11
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.07). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Midjourney Video](https://revana.si/models/midjourney-video.md): from $0.25 per generation
- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image

## Ideogram V3 API

> Best in class at text and graphic design: posters, logos, packaging and social graphics. By Ideogram, on Revana from $0.03 per image.

- Model id: `ideogram-v3`
- Kind: image
- Maker: Ideogram
- Tasks: Text to image, Image to image
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/ideogram-v3

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| balanced | $0.05 | 5 |
| quality | $0.07 | 7 |
| turbo | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-28)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | balanced | quality |
|---|---|---|
| Revana | $0.05 | $0.07 |
| Replicate | $0.06 | $0.09 |
| WaveSpeed | $0.06 | $0.09 |
| Runware | $0.06 | $0.09 |

Full comparison and sources: https://revana.si/blog/cheapest-ai-image-generation-api (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "ideogram-v3",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light"
  },
  "max_cost_usd": 0.08
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.05). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Ideogram Character](https://revana.si/models/ideogram-character.md): from $0.08 per image
- [Ideogram V4](https://revana.si/models/ideogram-v4.md): from $0.04 per image
- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image

## Qwen Image API

> Alibaba's open image model. Strong multilingual text rendering and instruction-based edits at a very low price. By Alibaba, on Revana $0.03 per image.

- Model id: `qwen-image`
- Kind: image
- Maker: Alibaba
- Tasks: Text to image, Image to image
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/qwen-image

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "qwen-image",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video
- [Wan 2.2 Animate](https://revana.si/models/wan-2.2-animate.md): from $0.08 per second of video

## HappyHorse 1.0 API

> Alibaba's HappyHorse 1.0: text to video, image to video, reference to video and video to video, up to 1080p. By Alibaba, on Revana from $0.17 per second of video.

- Model id: `happyhorse-1.0`
- Kind: video
- Maker: Alibaba
- Tasks: Text to video, Image to video, Reference to video, Video to video
- Resolutions: 720p, 1080p
- Durations: 3 to 15 seconds
- Audio: silent only
- Input images: up to 9
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/happyhorse-1.0

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.17 | 17 |
| 1080p | $0.29 | 29 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "happyhorse-1.0",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 3
  },
  "max_cost_usd": 0.77
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.51). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video
- [Wan 2.2 Animate](https://revana.si/models/wan-2.2-animate.md): from $0.08 per second of video

## HappyHorse 1.1 API

> Alibaba's HappyHorse 1.1: text to video, image to video and reference to video, up to 1080p. By Alibaba, on Revana from $0.15 per second of video.

- Model id: `happyhorse-1.1`
- Kind: video
- Maker: Alibaba
- Tasks: Text to video, Image to video, Reference to video
- Resolutions: 720p, 1080p
- Durations: 3 to 15 seconds
- Audio: silent only
- Input images: up to 9
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/happyhorse-1.1

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.15 | 15 |
| 1080p | $0.19 | 19 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 1080p |
|---|---|---|
| Revana | $0.15 | $0.19 |
| AI/ML API | $0.182 | $0.234 |

Full comparison and sources: https://revana.si/blog/ai-video-api-price-comparison (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "happyhorse-1.1",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 3
  },
  "max_cost_usd": 0.68
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.45). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [Wan 2.2 Animate](https://revana.si/models/wan-2.2-animate.md): from $0.08 per second of video

## Wan 2.2 Animate API

> Alibaba's Wan 2.2 Animate: video to video, up to 720p. By Alibaba, on Revana $0.08 per second of video.

- Model id: `wan-2.2-animate`
- Kind: video
- Maker: Alibaba
- Tasks: Video to video
- Resolutions: 720p
- Audio: silent only
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/wan-2.2-animate

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.08 | 8 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "wan-2.2-animate",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "video_url": "https://example.com/input.mp4",
    "resolution": "720p"
  },
  "max_cost_usd": 1.2
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.08 a second: $0.80 for a 10 s clip). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Wan 2.2 Turbo API

> Alibaba's Wan 2.2 Turbo: text to video and image to video, up to 720p. By Alibaba, on Revana from $0.05 per second of video.

- Model id: `wan-2.2-turbo`
- Kind: video
- Maker: Alibaba
- Tasks: Text to video, Image to video
- Resolutions: 480p, 720p
- Audio: silent only
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/wan-2.2-turbo

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.05 | 5 |
| 720p | $0.10 | 10 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "wan-2.2-turbo",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p"
  },
  "max_cost_usd": 1.5
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.10 a second: $1.00 for 10 s). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Wan 2.5 API

> Alibaba's Wan 2.5: text to video and image to video, up to 1080p, with sound. By Alibaba, on Revana from $0.05 per second of video.

- Model id: `wan-2.5`
- Kind: video
- Maker: Alibaba
- Tasks: Text to video, Image to video
- Resolutions: 480p, 720p, 1080p
- Durations: 5 s, 10 s
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4
- Audio: supported (`"audio": true`)
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/wan-2.5

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.05 | 5 |
| 720p | $0.08 | 8 |
| 1080p | $0.13 | 13 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "wan-2.5",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 5,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.6
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.40). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Wan 2.6 Flash API

> Alibaba's Wan 2.6 Flash: image to video and video to video, up to 1080p, with sound. By Alibaba, on Revana from $0.03 per second of video.

- Model id: `wan-2.6-flash`
- Kind: video
- Maker: Alibaba
- Tasks: Image to video, Video to video
- Resolutions: 720p, 1080p
- Durations: 5 s, 10 s, 15 s
- Audio: supported (`"audio": true`)
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/wan-2.6-flash

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.03 | 3 |
| 720p + with audio | $0.05 | 5 |
| 1080p | $0.04 | 4 |
| 1080p + with audio | $0.07 | 7 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "wan-2.6-flash",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_url": "https://example.com/start-frame.jpg",
    "resolution": "720p",
    "duration": 5,
    "audio": true
  },
  "max_cost_usd": 0.38
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.25). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Wan 2.6 I2V API

> Alibaba's Wan 2.6 I2V: image to video, up to 1080p. By Alibaba, on Revana from $0.09 per second of video. The lowest listed Wan 2.6 I2V price we found at 720p and 1080p (checked 2026-10-26).

- Model id: `wan-2.6-i2v`
- Kind: video
- Maker: Alibaba
- Tasks: Image to video
- Resolutions: 720p, 1080p
- Durations: 2 to 15 seconds
- Audio: silent only
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/wan-2.6-i2v

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.09 | 9 |
| 1080p | $0.14 | 14 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 1080p |
|---|---|---|
| Revana | $0.09 | $0.14 |
| OpenRouter | $0.10 | $0.15 |
| Vercel AI Gateway | $0.10 | $0.15 |

Revana has the lowest listed price per second of every API we checked at 720p ($0.09) and 1080p ($0.14).

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "wan-2.6-i2v",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_url": "https://example.com/start-frame.jpg",
    "resolution": "720p",
    "duration": 2
  },
  "max_cost_usd": 0.27
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.18). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Wan 2.6 I2V Flash API

> Alibaba's Wan 2.6 I2V Flash: image to video, up to 1080p, with sound. By Alibaba, on Revana from $0.03 per second of video.

- Model id: `wan-2.6-i2v-flash`
- Kind: video
- Maker: Alibaba
- Tasks: Image to video
- Resolutions: 720p, 1080p
- Durations: 2 to 15 seconds
- Audio: supported (`"audio": true`)
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/wan-2.6-i2v-flash

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.03 | 3 |
| 720p + with audio | $0.05 | 5 |
| 1080p | $0.04 | 4 |
| 1080p + with audio | $0.07 | 7 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "wan-2.6-i2v-flash",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_url": "https://example.com/start-frame.jpg",
    "resolution": "720p",
    "duration": 2,
    "audio": true
  },
  "max_cost_usd": 0.15
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.10). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Wan 2.7 API

> Alibaba's Wan 2.7: text to video, image to video, first and last frames, video to video and reference to video, up to 1080p. By Alibaba, on Revana from $0.09 per second of video.

- Model id: `wan-2.7`
- Kind: video
- Maker: Alibaba
- Tasks: Text to video, Image to video, First + last frame, Video to video, Reference to video
- Resolutions: 720p, 1080p
- Durations: 2 to 15 seconds
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4
- Audio: silent only
- Input images: up to 5
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/wan-2.7

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.09 | 9 |
| 1080p | $0.14 | 14 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 1080p |
|---|---|---|
| Revana | $0.09 | $0.14 |
| fal | $0.10 | $0.15 |
| OpenRouter | $0.10 | – |
| Vercel AI Gateway | $0.10 | $0.15 |

Full comparison and sources: https://revana.si/blog/wan-api-pricing (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "wan-2.7",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 2,
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.27
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.18). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Wan 2.7 R2V API

> Alibaba's Wan 2.7 R2V: reference to video, up to 1080p. By Alibaba, on Revana from $0.09 per second of video.

- Model id: `wan-2.7-r2v`
- Kind: video
- Maker: Alibaba
- Tasks: Reference to video
- Resolutions: 720p, 1080p
- Durations: 2 to 15 seconds
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4
- Audio: silent only
- Input images: up to 5
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/wan-2.7-r2v

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.09 | 9 |
| 1080p | $0.14 | 14 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "wan-2.7-r2v",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_urls": [
      "https://example.com/ref-1.jpg",
      "https://example.com/ref-2.jpg"
    ],
    "resolution": "720p",
    "duration": 2,
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.27
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.18). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Wan 2.7 Video Edit API

> Alibaba's Wan 2.7 Video Edit: video to video, up to 1080p. By Alibaba, on Revana from $0.09 per second of video.

- Model id: `wan-2.7-videoedit`
- Kind: video
- Maker: Alibaba
- Tasks: Video to video
- Resolutions: 720p, 1080p
- Durations: 2 to 10 seconds
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4
- Audio: silent only
- Input images: up to 4
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/wan-2.7-videoedit

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.09 | 9 |
| 1080p | $0.14 | 14 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "wan-2.7-videoedit",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "video_url": "https://example.com/input.mp4",
    "resolution": "720p",
    "duration": 2,
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 1.4
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.09 a second: $0.90 for a 10 s clip). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Flux 3 Video API

> Black Forest Labs's Flux 3 Video: text to video and image to video, up to fhd, with sound. By Black Forest Labs, on Revana from $0.06 per second of video.

- Model id: `flux-3-video`
- Kind: video
- Maker: Black Forest Labs
- Tasks: Text to video, Image to video
- Resolutions: draft, hd, fhd
- Durations: 5 to 20 seconds
- Audio: supported (`"audio": true`)
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/flux-3-video

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| draft | $0.06 | 6 |
| fhd | $0.29 | 29 |
| hd | $0.17 | 17 |
| v2v-draft | $0.12 | 12 |
| v2v-fhd | $0.53 | 53 |
| v2v-hd | $0.41 | 41 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "flux-3-video",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "draft",
    "duration": 5,
    "audio": true
  },
  "max_cost_usd": 0.45
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.30). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Flux 2 Pro](https://revana.si/models/flux-2-pro.md): from $0.03 per image
- [Flux 2 Flex](https://revana.si/models/flux-2-flex.md): from $0.05 per image
- [Flux 2 Max](https://revana.si/models/flux-2-max.md): from $0.07 per image
- [Flux 3](https://revana.si/models/flux-3.md): from $0.05 per image
- [Flux 3 Klein](https://revana.si/models/flux-3-klein.md): from $0.01 per image
- [Flux Kontext Max](https://revana.si/models/flux-kontext-max.md): from $0.07 per image

## ByteDance Lip Sync API

> ByteDance's ByteDance Lip Sync: video to video, with sound. By ByteDance, on Revana $0.05 per second of video.

- Model id: `bytedance-lipsync`
- Kind: video
- Maker: ByteDance
- Tasks: Video to video
- Audio: supported (`"audio": true`)
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/bytedance-lipsync

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| All settings | $0.05 | 5 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "bytedance-lipsync",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "video_url": "https://example.com/input.mp4",
    "audio_url": "https://example.com/voice.mp3"
  },
  "max_cost_usd": 0.75
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.05 a second: $0.50 for a 10 s clip). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image

## OmniHuman 1.5 API

> ByteDance's OmniHuman 1.5: image to video, with sound. By ByteDance, on Revana $0.17 per second of video.

- Model id: `omnihuman-1.5`
- Kind: video
- Maker: ByteDance
- Tasks: Image to video
- Audio: supported (`"audio": true`)
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/omnihuman-1.5

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| All settings | $0.17 | 17 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "omnihuman-1.5",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_url": "https://example.com/start-frame.jpg",
    "audio_url": "https://example.com/voice.mp3"
  },
  "max_cost_usd": 2.6
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.17 a second: $1.70 for a 10 s clip). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image

## OmniHuman 2.0 API

> ByteDance's newest talking-human model: one picture and a voice track (4 seconds to 5 minutes) become a 720p video of that person speaking, with natural lip sync, gestures and camera moves. By ByteDance, on Revana $0.09 per second of video.

- Model id: `omnihuman-2.0`
- Kind: video
- Maker: ByteDance
- Tasks: Image to video
- Audio: supported (`"audio": true`)
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/omnihuman-2.0

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| All settings | $0.09 | 9 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "omnihuman-2.0",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_url": "https://example.com/start-frame.jpg",
    "audio_url": "https://example.com/voice.mp3"
  },
  "max_cost_usd": 1.4
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.09 a second: $0.90 for a 10 s clip). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image

## Seedance 2.0 Face API

> ByteDance's Seedance 2.0 Face: text to video, image to video, first and last frames and reference to video, up to 1080p, with sound. By ByteDance, on Revana from $0.13 per second of video.

- Model id: `seedance-2.0-face`
- Kind: video
- Maker: ByteDance
- Tasks: Text to video, Image to video, First + last frame, Reference to video
- Resolutions: 480p, 720p, 1080p
- Durations: 4 to 15 seconds
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive
- Audio: supported (`"audio": true`)
- Input images: up to 9
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/seedance-2.0-face

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.13 | 13 |
| 720p | $0.27 | 27 |
| 1080p | $0.63 | 63 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "seedance-2.0-face",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 4,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 1.7
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $1.08). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image

## Seedance 2.0 Fast API

> ByteDance's Seedance 2.0 Fast: text to video, image to video, first and last frames and reference to video, up to 720p, with sound. By ByteDance, on Revana from $0.05 per second of video.

- Model id: `seedance-2.0-fast`
- Kind: video
- Maker: ByteDance
- Tasks: Text to video, Image to video, First + last frame, Reference to video
- Resolutions: 480p, 720p
- Durations: 4 to 15 seconds
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive
- Audio: supported (`"audio": true`)
- Input images: up to 9
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/seedance-2.0-fast

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.05 | 5 |
| 720p | $0.11 | 11 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p |
|---|---|---|
| Revana | $0.05 | $0.11 |
| BytePlus ModelArk | $0.0538 | $0.121 |
| fal | $0.1076 | $0.2419 |
| Atlas Cloud | $0.072 | $0.1548 |
| Replicate | $0.07 | $0.15 |
| Runway API | $0.29 | $0.29 |
| Kie.ai | $0.059 | $0.124 |
| WaveSpeed | $0.10 | $0.20 |
| Vercel AI Gateway | $0.0538 | $0.121 |
| Runware | $0.06 | $0.13 |
| AI/ML API | $0.07436 | $0.16003 |

Full comparison and sources: https://revana.si/blog/seedance-2-0-api-price-comparison (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "seedance-2.0-fast",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 4,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.66
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.44). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image

## Seedance 2.0 Fast Face API

> ByteDance's Seedance 2.0 Fast Face: text to video, image to video, first and last frames and reference to video, up to 720p, with sound. By ByteDance, on Revana from $0.10 per second of video.

- Model id: `seedance-2.0-fast-face`
- Kind: video
- Maker: ByteDance
- Tasks: Text to video, Image to video, First + last frame, Reference to video
- Resolutions: 480p, 720p
- Durations: 4 to 15 seconds
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive
- Audio: supported (`"audio": true`)
- Input images: up to 9
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/seedance-2.0-fast-face

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.10 | 10 |
| 720p | $0.22 | 22 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "seedance-2.0-fast-face",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 4,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 1.4
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.88). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image

## Seedance 2.0 Mini API

> ByteDance's Seedance 2.0 Mini: text to video, image to video, first and last frames and reference to video, up to 720p, with sound. By ByteDance, on Revana from $0.03 per second of video.

- Model id: `seedance-2.0-mini`
- Kind: video
- Maker: ByteDance
- Tasks: Text to video, Image to video, First + last frame, Reference to video
- Resolutions: 480p, 720p
- Durations: 4 to 15 seconds
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive
- Audio: supported (`"audio": true`)
- Input images: up to 9
- Routes: 4 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/seedance-2.0-mini

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.03 | 3 |
| 720p | $0.05 | 5 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p |
|---|---|---|
| Revana | $0.03 | $0.05 |
| BytePlus ModelArk | $0.0336 | $0.0756 |
| Atlas Cloud | $0.045 | $0.0968 |
| Replicate | $0.04 | $0.09 |
| Runway API | $0.16 | $0.16 |
| CometAPI | $0.032 | $0.064 |
| WaveSpeed | $0.06 | $0.12 |
| OpenRouter | $0.0336 | $0.0756 |
| Vercel AI Gateway | $0.0336 | $0.0756 |
| Runware | $0.036 | $0.081 |
| PiAPI | $0.042 | $0.084 |

Full comparison and sources: https://revana.si/blog/seedance-2-0-api-price-comparison (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "seedance-2.0-mini",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 4,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.3
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.20). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image

## Gemini Omni API

> Google's Gemini Omni: text to video and image to video, up to 4K, with sound. By Google, on Revana from $0.10 per second of video.

- Model id: `gemini-omni`
- Kind: video
- Maker: Google
- Tasks: Text to video, Image to video, Reference to video
- Resolutions: 720p, 1080p, 4K
- Durations: 4 s, 6 s, 8 s, 10 s
- Audio: supported (`"audio": true`)
- Input images: up to 7
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/gemini-omni

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.10 | 10 |
| 1080p | $0.10 | 10 |
| 4K | $0.23 | 23 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "gemini-omni",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 8,
    "audio": true
  },
  "max_cost_usd": 1.2
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.80). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video
- [Veo 3.1 Fast](https://revana.si/models/veo-3.1-fast.md): from $0.08 per second of video
- [Veo 3.1](https://revana.si/models/veo-3.1.md): from $0.20 per second of video
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image

## Gemini Omni 1.1 Flash API

> Google's Gemini Omni 1.1 Flash: text to video, image to video, first and last frames and reference to video, up to 4K, with sound. By Google, on Revana from $0.02 per second of video + $0.10 a job.

- Model id: `gemini-omni-1.1-flash`
- Kind: video
- Maker: Google
- Tasks: Text to video, Image to video, First + last frame, Reference to video
- Resolutions: 360p, 720p, 1080p, 4K
- Durations: 4 s, 6 s, 8 s, 10 s
- Aspect ratios: 16:9, 9:16
- Audio: supported (`"audio": true`)
- Input images: up to 7
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/gemini-omni-1.1-flash

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 360p | $0.02 + $0.10 a job | 2 + 10 |
| 720p | $0.03 + $0.18 a job | 3 + 18 |
| 1080p | $0.03 + $0.18 a job | 3 + 18 |
| 4K | $0.03 + $0.78 a job | 3 + 78 |

Each job is billed the per-job price once, plus the price a second for every second of video. 100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "gemini-omni-1.1-flash",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 8,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.63
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.42). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video
- [Veo 3.1 Fast](https://revana.si/models/veo-3.1-fast.md): from $0.08 per second of video
- [Veo 3.1](https://revana.si/models/veo-3.1.md): from $0.20 per second of video
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image

## Gemini Omni Flash Preview API

> Google's Gemini Omni Flash Preview: text to video, image to video and reference to video, up to 720p. By Google, on Revana $0.11 per second of video.

- Model id: `gemini-omni-flash-preview`
- Kind: video
- Maker: Google
- Tasks: Text to video, Image to video, Reference to video
- Resolutions: 720p
- Audio: silent only
- Input images: up to 16
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/gemini-omni-flash-preview

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| All settings | $0.11 | 11 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "gemini-omni-flash-preview",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p"
  },
  "max_cost_usd": 1.7
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.11 a second: $1.10 for 10 s). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video
- [Veo 3.1 Fast](https://revana.si/models/veo-3.1-fast.md): from $0.08 per second of video
- [Veo 3.1](https://revana.si/models/veo-3.1.md): from $0.20 per second of video
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image

## Veo 3.1 Lite API

> Google's Veo 3.1 Lite: text to video, image to video, first and last frames and reference to video, up to 4K, with sound. By Google, on Revana from $0.05 per second of video.

- Model id: `veo-3.1-lite`
- Kind: video
- Maker: Google
- Tasks: Text to video, Image to video, First + last frame, Reference to video
- Resolutions: 720p, 1080p, 4K
- Durations: 4 s, 6 s, 8 s
- Aspect ratios: 16:9, 9:16, Auto
- Audio: supported (`"audio": true`)
- Input images: up to 3
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/veo-3.1-lite

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.05 | 5 |
| 1080p | $0.06 | 6 |
| 4K | $0.24 | 24 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 720p | 1080p |
|---|---|---|
| Revana | $0.05 | $0.06 |
| Google's own API | – | $0.08 |
| fal | – | $0.08 |
| Replicate | – | $0.08 |
| WaveSpeed | – | $0.08 |
| OpenRouter | – | $0.08 |
| Vercel AI Gateway | – | $0.08 |
| Runware | – | $0.08 |
| AI/ML API | $0.065 | $0.104 |

Full comparison and sources: https://revana.si/blog/veo-3-1-api-pricing (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "veo-3.1-lite",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 8,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.6
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.40). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video
- [Veo 3.1 Fast](https://revana.si/models/veo-3.1-fast.md): from $0.08 per second of video
- [Veo 3.1](https://revana.si/models/veo-3.1.md): from $0.20 per second of video
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image

## Genjutsu Motion Transfer API

> Higgsfield's Genjutsu motion transfer: the people in your pictures take on every move of a 4 to 30 second clip, at 480p to 1080p. By Higgsfield, on Revana from $0.21 per second of video.

- Model id: `genjutsu-motion-transfer`
- Kind: video
- Maker: Higgsfield
- Tasks: Video to video
- Resolutions: 480p, 720p, 1080p
- Audio: silent only
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/genjutsu-motion-transfer

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.21 | 21 |
| 720p | $0.46 | 46 |
| 1080p | $1.12 | 112 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "genjutsu-motion-transfer",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "video_url": "https://example.com/input.mp4",
    "resolution": "720p"
  },
  "max_cost_usd": 6.9
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.46 a second: $4.60 for a 10 s clip). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image
- [Genjutsu Object Swap](https://revana.si/models/genjutsu-object-swap.md): from $0.21 per second of video
- [Marketing Studio Image](https://revana.si/models/marketing-studio-image.md): from $0.04 per image
- [Soul](https://revana.si/models/soul.md): from $0.13 per image
- [Soul Cinema](https://revana.si/models/soul-cinema.md): from $0.01 per image
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video

## Genjutsu Object Swap API

> Higgsfield's Genjutsu object swap: replace a person or object in a 4 to 30 second clip with the one in your pictures, keeping the motion, at 480p to 1080p. By Higgsfield, on Revana from $0.21 per second of video.

- Model id: `genjutsu-object-swap`
- Kind: video
- Maker: Higgsfield
- Tasks: Video to video
- Resolutions: 480p, 720p, 1080p
- Audio: silent only
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/genjutsu-object-swap

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.21 | 21 |
| 720p | $0.46 | 46 |
| 1080p | $1.12 | 112 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "genjutsu-object-swap",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "video_url": "https://example.com/input.mp4",
    "resolution": "720p"
  },
  "max_cost_usd": 6.9
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.46 a second: $4.60 for a 10 s clip). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image
- [Genjutsu Motion Transfer](https://revana.si/models/genjutsu-motion-transfer.md): from $0.21 per second of video
- [Marketing Studio Image](https://revana.si/models/marketing-studio-image.md): from $0.04 per image
- [Soul](https://revana.si/models/soul.md): from $0.13 per image
- [Soul Cinema](https://revana.si/models/soul-cinema.md): from $0.01 per image
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video

## Kling 2.1 Master API

> Kuaishou's Kling 2.1 Master: text to video and image to video. By Kuaishou, on Revana $0.32 per second of video.

- Model id: `kling-2.1-master`
- Kind: video
- Maker: Kuaishou
- Tasks: Text to video, Image to video
- Durations: 5 s, 10 s
- Audio: silent only
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/kling-2.1-master

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| standard | $0.32 | 32 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-2.1-master",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "duration": 5
  },
  "max_cost_usd": 2.4
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $1.60). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video
- [Kling 2.1 Standard](https://revana.si/models/kling-2.1-std.md): from $0.07 per second of video

## Kling 2.1 Pro API

> Kuaishou's Kling 2.1 Pro: image to video. By Kuaishou, on Revana $0.11 per second of video.

- Model id: `kling-2.1-pro`
- Kind: video
- Maker: Kuaishou
- Tasks: Image to video
- Durations: 5 s, 10 s
- Audio: silent only
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/kling-2.1-pro

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| standard | $0.11 | 11 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-2.1-pro",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_url": "https://example.com/start-frame.jpg",
    "duration": 5
  },
  "max_cost_usd": 0.83
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.55). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Standard](https://revana.si/models/kling-2.1-std.md): from $0.07 per second of video

## Kling 2.1 Standard API

> Kuaishou's Kling 2.1 Standard: image to video, up to 720p. By Kuaishou, on Revana $0.07 per second of video.

- Model id: `kling-2.1-std`
- Kind: video
- Maker: Kuaishou
- Tasks: Image to video
- Resolutions: 720p
- Durations: 5 s, 10 s
- Audio: silent only
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/kling-2.1-std

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.07 | 7 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-2.1-std",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_url": "https://example.com/start-frame.jpg",
    "resolution": "720p",
    "duration": 5
  },
  "max_cost_usd": 0.53
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.35). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video

## Kling 2.5 Turbo Pro API

> Kuaishou's Kling 2.5 Turbo Pro: text to video and image to video. By Kuaishou, on Revana $0.08 per second of video.

- Model id: `kling-2.5-turbo-pro`
- Kind: video
- Maker: Kuaishou
- Tasks: Text to video, Image to video
- Durations: 5 s, 10 s
- Audio: silent only
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/kling-2.5-turbo-pro

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| standard | $0.08 | 8 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-2.5-turbo-pro",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "duration": 5
  },
  "max_cost_usd": 0.6
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.40). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video

## Kling 2.6 Motion Control API

> Kuaishou's Kling 2.6 Motion Control: video to video, up to 1080p. By Kuaishou, on Revana from $0.08 per second of video.

- Model id: `kling-2.6-motion-control`
- Kind: video
- Maker: Kuaishou
- Tasks: Video to video
- Resolutions: 720p, 1080p
- Audio: silent only
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/kling-2.6-motion-control

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.08 | 8 |
| 1080p | $0.12 | 12 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-2.6-motion-control",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "video_url": "https://example.com/input.mp4",
    "resolution": "720p"
  },
  "max_cost_usd": 1.2
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.08 a second: $0.80 for a 10 s clip). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video

## Kling 2.6 Standard API

> Kuaishou's Kling 2.6 Standard: text to video and image to video, up to 720p. By Kuaishou, on Revana $0.05 per second of video.

- Model id: `kling-2.6-std`
- Kind: video
- Maker: Kuaishou
- Tasks: Text to video, Image to video
- Resolutions: 720p
- Durations: 5 s, 10 s
- Audio: silent only
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/kling-2.6-std

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| All settings | $0.05 | 5 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-2.6-std",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 5
  },
  "max_cost_usd": 0.38
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.25). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video

## Kling 3.0 Motion Control API

> Kuaishou's Kling 3.0 Motion Control: video to video, up to 1080p. By Kuaishou, on Revana from $0.13 per second of video.

- Model id: `kling-3.0-motion-control`
- Kind: video
- Maker: Kuaishou
- Tasks: Video to video
- Resolutions: 720p, 1080p
- Audio: silent only
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/kling-3.0-motion-control

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.13 | 13 |
| 1080p | $0.18 | 18 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-3.0-motion-control",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "video_url": "https://example.com/input.mp4",
    "resolution": "720p"
  },
  "max_cost_usd": 2
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.13 a second: $1.30 for a 10 s clip). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video

## Kling 3.0 Omni API

> Kuaishou's Kling 3.0 Omni: text to video, image to video, first and last frames, reference to video and video to video, up to 4K, with sound. By Kuaishou, on Revana from $0.09 per second of video.

- Model id: `kling-3.0-omni`
- Kind: video
- Maker: Kuaishou
- Tasks: Text to video, Image to video, First + last frame, Reference to video, Video to video
- Resolutions: 720p, 1080p, 4K
- Durations: 3 to 15 seconds
- Audio: supported (`"audio": true`)
- Input images: up to 7
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/kling-3.0-omni

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.09 | 9 |
| 720p + with audio | $0.13 | 13 |
| 720p + with video input | $0.13 | 13 |
| 1080p | $0.12 | 12 |
| 1080p + with audio | $0.14 | 14 |
| 1080p + with video input | $0.17 | 17 |
| 4K | $0.48 | 48 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-3.0-omni",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 3,
    "audio": true
  },
  "max_cost_usd": 0.59
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.39). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video

## Kling 3.0 Turbo API

> Kuaishou's Kling 3.0 Turbo: text to video and image to video, up to 1080p. By Kuaishou, on Revana from $0.13 per second of video.

- Model id: `kling-3.0-turbo`
- Kind: video
- Maker: Kuaishou
- Tasks: Text to video, Image to video
- Resolutions: 720p, 1080p
- Durations: 3 to 15 seconds
- Audio: silent only
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/kling-3.0-turbo

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.13 | 13 |
| 1080p | $0.16 | 16 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-3.0-turbo",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 3
  },
  "max_cost_usd": 0.59
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.39). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video

## Kling Avatar Pro API

> Kuaishou's Kling Avatar Pro: image to video, up to 1080p, with sound. By Kuaishou, on Revana $0.13 per second of video.

- Model id: `kling-avatar-pro`
- Kind: video
- Maker: Kuaishou
- Tasks: Image to video
- Resolutions: 1080p
- Durations: 1 to 300 seconds
- Audio: supported (`"audio": true`)
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/kling-avatar-pro

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| All settings | $0.13 | 13 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-avatar-pro",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_url": "https://example.com/start-frame.jpg",
    "audio_url": "https://example.com/voice.mp3",
    "resolution": "1080p",
    "duration": 1
  },
  "max_cost_usd": 2
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.13 a second: $1.30 for a 10 s clip). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video

## Kling Avatar Standard API

> Kuaishou's Kling Avatar Standard: image to video, up to 720p, with sound. By Kuaishou, on Revana $0.07 per second of video.

- Model id: `kling-avatar-std`
- Kind: video
- Maker: Kuaishou
- Tasks: Image to video
- Resolutions: 720p
- Durations: 1 to 300 seconds
- Audio: supported (`"audio": true`)
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/kling-avatar-std

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| All settings | $0.07 | 7 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-avatar-std",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_url": "https://example.com/start-frame.jpg",
    "audio_url": "https://example.com/voice.mp3",
    "resolution": "720p",
    "duration": 1
  },
  "max_cost_usd": 1.1
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.07 a second: $0.70 for a 10 s clip). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video

## Kling O1 API

> Kuaishou's Kling O1: text to video, image to video and reference to video. By Kuaishou, on Revana from $0.09 per second of video.

- Model id: `kling-o1`
- Kind: video
- Maker: Kuaishou
- Tasks: Text to video, Image to video, Reference to video
- Durations: 5 s, 10 s
- Audio: silent only
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/kling-o1

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.09 | 9 |
| 1080p | $0.12 | 12 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "kling-o1",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "duration": 5
  },
  "max_cost_usd": 0.68
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.45). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Kling 4.0](https://revana.si/models/kling-4.0.md): from $0.135 per second of video
- [Kling 4.0 Flash](https://revana.si/models/kling-4.0-flash.md): from $0.10 per second of video
- [Kling 3.0](https://revana.si/models/kling-3.0.md): from $0.09 per second of video
- [Kling 2.6 Pro](https://revana.si/models/kling-2.6-pro.md): from $0.08 per second of video
- [Kling 2.1 Master](https://revana.si/models/kling-2.1-master.md): from $0.32 per second of video
- [Kling 2.1 Pro](https://revana.si/models/kling-2.1-pro.md): from $0.11 per second of video

## LTX 2.5 Fast API

> Lightricks's LTX 2.5 Fast: text to video, image to video and first and last frames, up to 4K. By Lightricks, on Revana from $0.12 per second of video.

- Model id: `ltx-2.5-fast`
- Kind: video
- Maker: Lightricks
- Tasks: Text to video, Image to video, First + last frame
- Resolutions: 720p, 1080p, 2K, 4K
- Durations: 6 s, 8 s, 10 s
- Audio: silent only
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/ltx-2.5-fast

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.12 | 12 |
| 1080p | $0.18 | 18 |
| 2K | $0.26 | 26 |
| 4K | $0.40 | 40 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "ltx-2.5-fast",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 8
  },
  "max_cost_usd": 1.5
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.96). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [LTX 2.5 Pro](https://revana.si/models/ltx-2.5-pro.md): from $0.16 per second of video
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Nemesis (aka Genjutsu)](https://revana.si/models/nemesis.md): from $0.08 per second of video

## LTX 2.5 Pro API

> Lightricks's LTX 2.5 Pro: text to video, image to video and first and last frames, up to 1080p. By Lightricks, on Revana from $0.16 per second of video.

- Model id: `ltx-2.5-pro`
- Kind: video
- Maker: Lightricks
- Tasks: Text to video, Image to video, First + last frame
- Resolutions: 720p, 1080p
- Durations: 6 s, 8 s, 10 s
- Audio: silent only
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/ltx-2.5-pro

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.16 | 16 |
| 1080p | $0.23 | 23 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "ltx-2.5-pro",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 8
  },
  "max_cost_usd": 2
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $1.28). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [LTX 2.5 Fast](https://revana.si/models/ltx-2.5-fast.md): from $0.12 per second of video
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Nemesis (aka Genjutsu)](https://revana.si/models/nemesis.md): from $0.08 per second of video

## Midjourney Video API

> Midjourney's Midjourney Video: image to video, up to 720p. By Midjourney, on Revana from $0.25 per generation.

- Model id: `midjourney-video`
- Kind: video
- Maker: Midjourney
- Tasks: Image to video
- Resolutions: 480p, 720p
- Audio: silent only
- Routes: 1
- Status: coming soon; requests are refused and not charged
- Page: https://revana.si/models/midjourney-video

### Price (per generation, USD)

| Setting | Price | Credits |
|---|---|---|
| 480p | $0.25 | 25 |
| 720p | $0.50 | 50 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Similar models

- [Midjourney V7](https://revana.si/models/midjourney-v7.md): from $0.06 per image
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Nemesis (aka Genjutsu)](https://revana.si/models/nemesis.md): from $0.08 per second of video

## Hailuo 02 API

> MiniMax's Hailuo 02: text to video, image to video and first and last frames, up to 1080p. By MiniMax, on Revana from $0.02 per second of video.

- Model id: `hailuo-02`
- Kind: video
- Maker: MiniMax
- Tasks: Text to video, Image to video, First + last frame
- Resolutions: 512p, 768p, 1080p
- Durations: 5 s, 10 s
- Audio: silent only
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/hailuo-02

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 512p | $0.02 | 2 |
| 768p | $0.04 | 4 |
| 1080p | $0.08 | 8 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "hailuo-02",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "1080p",
    "duration": 5
  },
  "max_cost_usd": 0.6
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.40). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Hailuo 2.3](https://revana.si/models/hailuo-2.3.md): from $0.06 per second of video
- [Hailuo 02 Pro](https://revana.si/models/hailuo-02-pro.md): from $0.06 per second of video
- [Hailuo 02 Standard](https://revana.si/models/hailuo-02-std.md): from $0.02 per second of video
- [Hailuo 2.3 Fast](https://revana.si/models/hailuo-2.3-fast.md): from $0.04 per second of video
- [Hailuo 2.3 Standard](https://revana.si/models/hailuo-2.3-std.md): from $0.04 per second of video
- [MiniMax H3](https://revana.si/models/minimax-h3.md): from $0.05 per second of video

## Hailuo 02 Pro API

> MiniMax's Hailuo 02 Pro: text to video and image to video, up to 1080p. By MiniMax, on Revana $0.06 per second of video.

- Model id: `hailuo-02-pro`
- Kind: video
- Maker: MiniMax
- Tasks: Text to video, Image to video
- Resolutions: 1080p
- Durations: 6 s
- Audio: silent only
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/hailuo-02-pro

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| All settings | $0.06 | 6 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "hailuo-02-pro",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "1080p",
    "duration": 6
  },
  "max_cost_usd": 0.54
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.36). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Hailuo 2.3](https://revana.si/models/hailuo-2.3.md): from $0.06 per second of video
- [Hailuo 02](https://revana.si/models/hailuo-02.md): from $0.02 per second of video
- [Hailuo 02 Standard](https://revana.si/models/hailuo-02-std.md): from $0.02 per second of video
- [Hailuo 2.3 Fast](https://revana.si/models/hailuo-2.3-fast.md): from $0.04 per second of video
- [Hailuo 2.3 Standard](https://revana.si/models/hailuo-2.3-std.md): from $0.04 per second of video
- [MiniMax H3](https://revana.si/models/minimax-h3.md): from $0.05 per second of video

## Hailuo 02 Standard API

> MiniMax's Hailuo 02 Standard: text to video and image to video, up to 768p. By MiniMax, on Revana from $0.02 per second of video.

- Model id: `hailuo-02-std`
- Kind: video
- Maker: MiniMax
- Tasks: Text to video, Image to video
- Resolutions: 512p, 768p
- Durations: 6 s, 10 s
- Audio: silent only
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/hailuo-02-std

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 512p | $0.02 | 2 |
| 768p | $0.04 | 4 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "hailuo-02-std",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "512p",
    "duration": 6
  },
  "max_cost_usd": 0.18
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.12). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Hailuo 2.3](https://revana.si/models/hailuo-2.3.md): from $0.06 per second of video
- [Hailuo 02](https://revana.si/models/hailuo-02.md): from $0.02 per second of video
- [Hailuo 02 Pro](https://revana.si/models/hailuo-02-pro.md): from $0.06 per second of video
- [Hailuo 2.3 Fast](https://revana.si/models/hailuo-2.3-fast.md): from $0.04 per second of video
- [Hailuo 2.3 Standard](https://revana.si/models/hailuo-2.3-std.md): from $0.04 per second of video
- [MiniMax H3](https://revana.si/models/minimax-h3.md): from $0.05 per second of video

## Hailuo 2.3 Fast API

> MiniMax's Hailuo 2.3 Fast: image to video, up to 1080p. By MiniMax, on Revana from $0.04 per second of video.

- Model id: `hailuo-2.3-fast`
- Kind: video
- Maker: MiniMax
- Tasks: Image to video
- Resolutions: 768p, 1080p
- Durations: 6 s, 10 s
- Audio: silent only
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/hailuo-2.3-fast

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 768p | $0.04 | 4 |
| 1080p | $0.06 | 6 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "hailuo-2.3-fast",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_url": "https://example.com/start-frame.jpg",
    "resolution": "1080p",
    "duration": 6
  },
  "max_cost_usd": 0.54
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.36). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Hailuo 2.3](https://revana.si/models/hailuo-2.3.md): from $0.06 per second of video
- [Hailuo 02](https://revana.si/models/hailuo-02.md): from $0.02 per second of video
- [Hailuo 02 Pro](https://revana.si/models/hailuo-02-pro.md): from $0.06 per second of video
- [Hailuo 02 Standard](https://revana.si/models/hailuo-02-std.md): from $0.02 per second of video
- [Hailuo 2.3 Standard](https://revana.si/models/hailuo-2.3-std.md): from $0.04 per second of video
- [MiniMax H3](https://revana.si/models/minimax-h3.md): from $0.05 per second of video

## Hailuo 2.3 Standard API

> MiniMax's Hailuo 2.3 Standard: image to video, up to 1080p. By MiniMax, on Revana from $0.04 per second of video.

- Model id: `hailuo-2.3-std`
- Kind: video
- Maker: MiniMax
- Tasks: Image to video
- Resolutions: 768p, 1080p
- Durations: 6 s, 10 s
- Audio: silent only
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/hailuo-2.3-std

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 768p | $0.04 | 4 |
| 1080p | $0.06 | 6 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "hailuo-2.3-std",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_url": "https://example.com/start-frame.jpg",
    "resolution": "1080p",
    "duration": 6
  },
  "max_cost_usd": 0.54
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.36). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Hailuo 2.3](https://revana.si/models/hailuo-2.3.md): from $0.06 per second of video
- [Hailuo 02](https://revana.si/models/hailuo-02.md): from $0.02 per second of video
- [Hailuo 02 Pro](https://revana.si/models/hailuo-02-pro.md): from $0.06 per second of video
- [Hailuo 02 Standard](https://revana.si/models/hailuo-02-std.md): from $0.02 per second of video
- [Hailuo 2.3 Fast](https://revana.si/models/hailuo-2.3-fast.md): from $0.04 per second of video
- [MiniMax H3](https://revana.si/models/minimax-h3.md): from $0.05 per second of video

## MiniMax H3 API

> MiniMax's MiniMax H3: text to video, image to video, first and last frames and reference to video, up to 2K, with sound. By MiniMax, on Revana from $0.05 per second of video.

- Model id: `minimax-h3`
- Kind: video
- Maker: MiniMax
- Tasks: Text to video, Image to video, First + last frame, Reference to video
- Resolutions: 768p, 2K
- Durations: 4 to 15 seconds
- Aspect ratios: 21:9, 16:9, 4:3, 1:1, 3:4, 9:16, adaptive
- Audio: supported (`"audio": true`)
- Input images: up to 9
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/minimax-h3

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 768p | $0.05 | 5 |
| 768p + with video input | $0.05 | 5 |
| 2K | $0.08 | 8 |
| 2K + with video input | $0.08 | 8 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 768p | 2K |
|---|---|---|
| Revana | $0.05 | $0.08 |
| Vercel AI Gateway | $0.08 | $0.13 |


### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "minimax-h3",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "768p",
    "duration": 4,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.3
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.20). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Hailuo 2.3](https://revana.si/models/hailuo-2.3.md): from $0.06 per second of video
- [Hailuo 02](https://revana.si/models/hailuo-02.md): from $0.02 per second of video
- [Hailuo 02 Pro](https://revana.si/models/hailuo-02-pro.md): from $0.06 per second of video
- [Hailuo 02 Standard](https://revana.si/models/hailuo-02-std.md): from $0.02 per second of video
- [Hailuo 2.3 Fast](https://revana.si/models/hailuo-2.3-fast.md): from $0.04 per second of video
- [Hailuo 2.3 Standard](https://revana.si/models/hailuo-2.3-std.md): from $0.04 per second of video

## MiniMax H3 Max API

> MiniMax's MiniMax H3 Max: text to video, image to video and reference to video, up to 1080p. By MiniMax, on Revana from $0.05 per second of video.

- Model id: `minimax-h3-max`
- Kind: video
- Maker: MiniMax
- Tasks: Text to video, Image to video, Reference to video
- Resolutions: 480p, 768p, 1080p
- Durations: 5 to 15 seconds
- Audio: silent only
- Input images: up to 9
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/minimax-h3-max

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.05 | 5 |
| 768p | $0.08 | 8 |
| 1080p | $0.16 | 16 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "minimax-h3-max",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "1080p",
    "duration": 5
  },
  "max_cost_usd": 1.2
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.80). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Hailuo 2.3](https://revana.si/models/hailuo-2.3.md): from $0.06 per second of video
- [Hailuo 02](https://revana.si/models/hailuo-02.md): from $0.02 per second of video
- [Hailuo 02 Pro](https://revana.si/models/hailuo-02-pro.md): from $0.06 per second of video
- [Hailuo 02 Standard](https://revana.si/models/hailuo-02-std.md): from $0.02 per second of video
- [Hailuo 2.3 Fast](https://revana.si/models/hailuo-2.3-fast.md): from $0.04 per second of video
- [Hailuo 2.3 Standard](https://revana.si/models/hailuo-2.3-std.md): from $0.04 per second of video

## MiniMax H3 Regeneration API

> MiniMax's MiniMax H3 Regeneration: upscaling, up to 2K. By MiniMax, on Revana $0.05 per second of video.

- Model id: `minimax-h3-regeneration`
- Kind: video
- Maker: MiniMax
- Tasks: Upscale
- Resolutions: 2K
- Audio: silent only
- Routes: 1
- Status: coming soon; requests are refused and not charged
- Page: https://revana.si/models/minimax-h3-regeneration

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 2K | $0.05 | 5 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Similar models

- [Hailuo 2.3](https://revana.si/models/hailuo-2.3.md): from $0.06 per second of video
- [Hailuo 02](https://revana.si/models/hailuo-02.md): from $0.02 per second of video
- [Hailuo 02 Pro](https://revana.si/models/hailuo-02-pro.md): from $0.06 per second of video
- [Hailuo 02 Standard](https://revana.si/models/hailuo-02-std.md): from $0.02 per second of video
- [Hailuo 2.3 Fast](https://revana.si/models/hailuo-2.3-fast.md): from $0.04 per second of video
- [Hailuo 2.3 Standard](https://revana.si/models/hailuo-2.3-std.md): from $0.04 per second of video

## PixVerse V6 API

> PixVerse's PixVerse V6: text to video, image to video, first and last frames, reference to video and video to video, up to 1080p, with sound. By PixVerse, on Revana from $0.02 per second of video.

- Model id: `pixverse-v6`
- Kind: video
- Maker: PixVerse
- Tasks: Text to video, Image to video, First + last frame, Reference to video, Video to video
- Resolutions: 360p, 540p, 720p, 1080p
- Durations: 1 to 15 seconds
- Aspect ratios: 16:9, 4:3, 1:1, 3:4, 9:16, 2:3, 3:2, 21:9
- Audio: supported (`"audio": true`)
- Input images: up to 7
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/pixverse-v6

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 360p | $0.02 | 2 |
| 360p + with audio | $0.03 | 3 |
| 540p | $0.03 | 3 |
| 540p + with audio | $0.04 | 4 |
| 720p | $0.04 | 4 |
| 720p + with audio | $0.05 | 5 |
| 1080p | $0.08 | 8 |
| 1080p + with audio | $0.10 | 10 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-28)

Per second of output for a 16:9 clip; the +audio columns are with sound. Card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 360p | 360p+audio | 540p | 540p+audio | 720p | 720p+audio | 1080p | 1080p+audio |
|---|---|---|---|---|---|---|---|---|
| Revana | $0.02 | $0.03 | $0.03 | $0.04 | $0.04 | $0.05 | $0.08 | $0.10 |
| fal | $0.025 | $0.035 | $0.035 | $0.045 | $0.045 | $0.06 | $0.09 | $0.115 |
| Replicate | $0.05 | $0.07 | $0.07 | $0.09 | $0.09 | $0.12 | $0.18 | $0.23 |
| WaveSpeed | $0.025 | $0.035 | $0.035 | $0.045 | $0.045 | $0.06 | $0.09 | $0.115 |
| Runware | $0.025 | $0.035 | $0.035 | $0.045 | $0.045 | $0.06 | $0.09 | $0.115 |

Full comparison and sources: https://revana.si/blog/pixverse-v6-api-pricing (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "pixverse-v6",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 1,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.08
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.05). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [PixVerse V6 Max](https://revana.si/models/pixverse-v6-max.md): from $0.03 per second of video
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Nemesis (aka Genjutsu)](https://revana.si/models/nemesis.md): from $0.08 per second of video

## PixVerse V6 Max API

> PixVerse's PixVerse V6 Max: text to video and image to video, up to 1080p, with sound. By PixVerse, on Revana from $0.03 per second of video.

- Model id: `pixverse-v6-max`
- Kind: video
- Maker: PixVerse
- Tasks: Text to video, Image to video
- Resolutions: 360p, 540p, 720p, 1080p
- Durations: 1 to 15 seconds
- Aspect ratios: 16:9, 9:16, 4:3, 3:4, 1:1, 2:3, 3:2, 21:9
- Audio: supported (`"audio": true`)
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/pixverse-v6-max

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 360p | $0.03 | 3 |
| 360p + with audio | $0.04 | 4 |
| 540p | $0.04 | 4 |
| 540p + with audio | $0.05 | 5 |
| 720p | $0.05 | 5 |
| 720p + with audio | $0.07 | 7 |
| 1080p | $0.10 | 10 |
| 1080p + with audio | $0.13 | 13 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "pixverse-v6-max",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 1,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.11
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.07). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [PixVerse V6](https://revana.si/models/pixverse-v6.md): from $0.02 per second of video
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Nemesis (aka Genjutsu)](https://revana.si/models/nemesis.md): from $0.08 per second of video

## Runway API

> Runway's Runway: text to video, image to video and video to video, up to 1080p. By Runway, on Revana from $0.02 per second of video.

- Model id: `runway`
- Kind: video
- Maker: Runway
- Tasks: Text to video, Image to video, Video to video
- Resolutions: 720p, 1080p
- Durations: 5 s, 10 s
- Audio: silent only
- Routes: 1
- Status: coming soon; requests are refused and not charged
- Page: https://revana.si/models/runway

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.02 | 2 |
| 1080p | $0.04 | 4 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Nemesis (aka Genjutsu)](https://revana.si/models/nemesis.md): from $0.08 per second of video
- [Nemesis Lite](https://revana.si/models/nemesis-lite.md): from $0.06 per second of video

## Vidu Q3 API

> Shengshu's Vidu Q3: reference to video, up to 1080p. By Shengshu, on Revana from $0.05 per second of video.

- Model id: `vidu-q3`
- Kind: video
- Maker: Shengshu
- Tasks: Reference to video
- Resolutions: 540p, 720p, 1080p
- Durations: 3 to 16 seconds
- Aspect ratios: 16:9, 9:16, 4:3, 3:4, 1:1
- Audio: silent only
- Input images: up to 7
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/vidu-q3

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 540p | $0.05 | 5 |
| 720p | $0.10 | 10 |
| 1080p | $0.13 | 13 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "vidu-q3",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_urls": [
      "https://example.com/ref-1.jpg",
      "https://example.com/ref-2.jpg"
    ],
    "resolution": "720p",
    "duration": 3,
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.45
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.30). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Vidu Q3 Mix](https://revana.si/models/vidu-q3-mix.md): from $0.13 per second of video
- [Vidu Q3 Pro](https://revana.si/models/vidu-q3-pro.md): from $0.07 per second of video
- [Vidu Q3 Turbo](https://revana.si/models/vidu-q3-turbo.md): from $0.04 per second of video
- [Vidu Q4](https://revana.si/models/vidu-q4.md): from $0.05 per second of video
- [Vidu Q4 Preview](https://revana.si/models/vidu-q4-preview.md): from $0.06 per second of video
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video

## Vidu Q3 Mix API

> Shengshu's Vidu Q3 Mix: reference to video, up to 1080p. By Shengshu, on Revana from $0.13 per second of video.

- Model id: `vidu-q3-mix`
- Kind: video
- Maker: Shengshu
- Tasks: Reference to video
- Resolutions: 720p, 1080p
- Durations: 1 to 16 seconds
- Aspect ratios: 16:9, 9:16, 4:3, 3:4, 1:1
- Audio: silent only
- Input images: up to 7
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/vidu-q3-mix

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 720p | $0.13 | 13 |
| 1080p | $0.15 | 15 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "vidu-q3-mix",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_urls": [
      "https://example.com/ref-1.jpg",
      "https://example.com/ref-2.jpg"
    ],
    "resolution": "720p",
    "duration": 1,
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.2
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.13). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Vidu Q3](https://revana.si/models/vidu-q3.md): from $0.05 per second of video
- [Vidu Q3 Pro](https://revana.si/models/vidu-q3-pro.md): from $0.07 per second of video
- [Vidu Q3 Turbo](https://revana.si/models/vidu-q3-turbo.md): from $0.04 per second of video
- [Vidu Q4](https://revana.si/models/vidu-q4.md): from $0.05 per second of video
- [Vidu Q4 Preview](https://revana.si/models/vidu-q4-preview.md): from $0.06 per second of video
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video

## Vidu Q3 Pro API

> Shengshu's Vidu Q3 Pro: text to video, image to video and first and last frames, up to 1080p, with sound. By Shengshu, on Revana from $0.07 per second of video.

- Model id: `vidu-q3-pro`
- Kind: video
- Maker: Shengshu
- Tasks: Text to video, Image to video, First + last frame
- Resolutions: 540p, 720p, 1080p
- Durations: 1 to 16 seconds
- Aspect ratios: 16:9, 9:16, 4:3, 3:4, 1:1
- Audio: supported (`"audio": true`)
- Input images: up to 2
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/vidu-q3-pro

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 540p | $0.07 | 7 |
| 720p | $0.15 | 15 |
| 1080p | $0.16 | 16 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "vidu-q3-pro",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 1,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.23
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.15). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Vidu Q3](https://revana.si/models/vidu-q3.md): from $0.05 per second of video
- [Vidu Q3 Mix](https://revana.si/models/vidu-q3-mix.md): from $0.13 per second of video
- [Vidu Q3 Turbo](https://revana.si/models/vidu-q3-turbo.md): from $0.04 per second of video
- [Vidu Q4](https://revana.si/models/vidu-q4.md): from $0.05 per second of video
- [Vidu Q4 Preview](https://revana.si/models/vidu-q4-preview.md): from $0.06 per second of video
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video

## Vidu Q3 Turbo API

> Shengshu's Vidu Q3 Turbo: text to video, image to video and first and last frames, up to 1080p, with sound. By Shengshu, on Revana from $0.04 per second of video.

- Model id: `vidu-q3-turbo`
- Kind: video
- Maker: Shengshu
- Tasks: Text to video, Image to video, First + last frame
- Resolutions: 540p, 720p, 1080p
- Durations: 1 to 16 seconds
- Aspect ratios: 16:9, 9:16, 4:3, 3:4, 1:1
- Audio: supported (`"audio": true`)
- Input images: up to 2
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/vidu-q3-turbo

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 540p | $0.04 | 4 |
| 720p | $0.06 | 6 |
| 1080p | $0.07 | 7 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "vidu-q3-turbo",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 1,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.09
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.06). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Vidu Q3](https://revana.si/models/vidu-q3.md): from $0.05 per second of video
- [Vidu Q3 Mix](https://revana.si/models/vidu-q3-mix.md): from $0.13 per second of video
- [Vidu Q3 Pro](https://revana.si/models/vidu-q3-pro.md): from $0.07 per second of video
- [Vidu Q4](https://revana.si/models/vidu-q4.md): from $0.05 per second of video
- [Vidu Q4 Preview](https://revana.si/models/vidu-q4-preview.md): from $0.06 per second of video
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video

## Vidu Q4 API

> Shengshu's Vidu Q4: image to video and reference to video, up to 4K. By Shengshu, on Revana from $0.05 per second of video.

- Model id: `vidu-q4`
- Kind: video
- Maker: Shengshu
- Tasks: Image to video, Reference to video
- Resolutions: 540p, 720p, 1080p, 2K, 4K
- Audio: silent only
- Input images: up to 15
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/vidu-q4

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 540p | $0.05 | 5 |
| 720p | $0.10 | 10 |
| 1080p | $0.12 | 12 |
| 2K | $0.19 | 19 |
| 4K | $0.39 | 39 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "vidu-q4",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_url": "https://example.com/start-frame.jpg",
    "resolution": "720p"
  },
  "max_cost_usd": 1.5
}
```

`max_cost_usd` is your cap, the most this request may cost ($0.10 a second: $1.00 for 10 s). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Vidu Q3](https://revana.si/models/vidu-q3.md): from $0.05 per second of video
- [Vidu Q3 Mix](https://revana.si/models/vidu-q3-mix.md): from $0.13 per second of video
- [Vidu Q3 Pro](https://revana.si/models/vidu-q3-pro.md): from $0.07 per second of video
- [Vidu Q3 Turbo](https://revana.si/models/vidu-q3-turbo.md): from $0.04 per second of video
- [Vidu Q4 Preview](https://revana.si/models/vidu-q4-preview.md): from $0.06 per second of video
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video

## Vidu Q4 Preview API

> Shengshu's Vidu Q4 Preview: image to video and reference to video, up to 4K, with sound. By Shengshu, on Revana from $0.06 per second of video.

- Model id: `vidu-q4-preview`
- Kind: video
- Maker: Shengshu
- Tasks: Image to video, Reference to video
- Resolutions: 540p, 720p, 1080p, 2K, 4K
- Durations: 3 to 16 seconds
- Aspect ratios: 16:9, 9:16, 4:3, 3:4, 1:1
- Audio: supported (`"audio": true`)
- Input images: up to 15
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/vidu-q4-preview

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 540p | $0.06 | 6 |
| 720p | $0.12 | 12 |
| 1080p | $0.13 | 13 |
| 2K | $0.21 | 21 |
| 4K | $0.42 | 42 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "vidu-q4-preview",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "image_url": "https://example.com/start-frame.jpg",
    "resolution": "720p",
    "duration": 3,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.54
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.36). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Vidu Q3](https://revana.si/models/vidu-q3.md): from $0.05 per second of video
- [Vidu Q3 Mix](https://revana.si/models/vidu-q3-mix.md): from $0.13 per second of video
- [Vidu Q3 Pro](https://revana.si/models/vidu-q3-pro.md): from $0.07 per second of video
- [Vidu Q3 Turbo](https://revana.si/models/vidu-q3-turbo.md): from $0.04 per second of video
- [Vidu Q4](https://revana.si/models/vidu-q4.md): from $0.05 per second of video
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video

## SkyReels V4 Fast API

> Skywork's SkyReels V4 Fast: text to video, image to video, first and last frames and reference to video, up to 1080p. By Skywork, on Revana from $0.08 per second of video.

- Model id: `skyreels-v4-fast`
- Kind: video
- Maker: Skywork
- Tasks: Text to video, Image to video, First + last frame, Reference to video
- Resolutions: 480p, 720p, 1080p
- Durations: 3 to 15 seconds
- Aspect ratios: 16:9, 4:3, 1:1, 9:16, 3:4
- Audio: silent only
- Input images: up to 5
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/skyreels-v4-fast

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.08 | 8 |
| 720p | $0.11 | 11 |
| 1080p | $0.28 | 28 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "skyreels-v4-fast",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 3,
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.5
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.33). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [SkyReels V4 Standard](https://revana.si/models/skyreels-v4-std.md): from $0.11 per second of video
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Nemesis (aka Genjutsu)](https://revana.si/models/nemesis.md): from $0.08 per second of video

## SkyReels V4 Standard API

> Skywork's SkyReels V4 Standard: text to video, image to video, first and last frames and reference to video, up to 1080p. By Skywork, on Revana from $0.11 per second of video.

- Model id: `skyreels-v4-std`
- Kind: video
- Maker: Skywork
- Tasks: Text to video, Image to video, First + last frame, Reference to video
- Resolutions: 480p, 720p, 1080p
- Durations: 3 to 15 seconds
- Aspect ratios: 16:9, 4:3, 1:1, 9:16, 3:4
- Audio: silent only
- Input images: up to 5
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/skyreels-v4-std

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.11 | 11 |
| 720p | $0.14 | 14 |
| 1080p | $0.35 | 35 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "skyreels-v4-std",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 3,
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.63
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.42). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [SkyReels V4 Fast](https://revana.si/models/skyreels-v4-fast.md): from $0.08 per second of video
- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Nemesis (aka Genjutsu)](https://revana.si/models/nemesis.md): from $0.08 per second of video

## Grok Imagine Extend API

> xAI's Grok Imagine Extend: video to video, up to 720p. By xAI, on Revana from $0.02 per second of video.

- Model id: `grok-imagine-extend`
- Kind: video
- Maker: xAI
- Tasks: Video to video
- Resolutions: 480p, 720p
- Durations: 6 s, 10 s
- Audio: silent only
- Routes: 1
- Status: coming soon; requests are refused and not charged
- Page: https://revana.si/models/grok-imagine-extend

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.02 | 2 |
| 720p | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Similar models

- [Grok Imagine Upscale](https://revana.si/models/grok-imagine-upscale.md): from $0.07 per generation
- [Grok Imagine Video](https://revana.si/models/grok-imagine-video.md): from $0.02 per second of video
- [Grok Imagine Video 1.5](https://revana.si/models/grok-imagine-video-1.5.md): from $0.04 per second of video
- [Grok Imagine Video 1.5 Preview](https://revana.si/models/grok-imagine-video-1.5-preview.md): from $0.02 per second of video
- [Grok Imagine Image](https://revana.si/models/grok-imagine-image.md): from $0.02 per image
- [Grok Imagine Image 2.0](https://revana.si/models/grok-imagine-image-2.0.md): from $0.02 per image

## Grok Imagine Upscale API

> xAI's Grok Imagine Upscale: upscaling, up to 1080p. By xAI, on Revana from $0.07 per generation.

- Model id: `grok-imagine-upscale`
- Kind: video
- Maker: xAI
- Tasks: Upscale
- Resolutions: 720p, 1080p
- Audio: silent only
- Routes: 1
- Status: coming soon; requests are refused and not charged
- Page: https://revana.si/models/grok-imagine-upscale

### Price (per generation, USD)

| Setting | Price | Credits |
|---|---|---|
| 480p->1080p | $0.19 | 19 |
| 480p->720p | $0.07 | 7 |
| 720p->1080p | $0.13 | 13 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Similar models

- [Grok Imagine Extend](https://revana.si/models/grok-imagine-extend.md): from $0.02 per second of video
- [Grok Imagine Video](https://revana.si/models/grok-imagine-video.md): from $0.02 per second of video
- [Grok Imagine Video 1.5](https://revana.si/models/grok-imagine-video-1.5.md): from $0.04 per second of video
- [Grok Imagine Video 1.5 Preview](https://revana.si/models/grok-imagine-video-1.5-preview.md): from $0.02 per second of video
- [Grok Imagine Image](https://revana.si/models/grok-imagine-image.md): from $0.02 per image
- [Grok Imagine Image 2.0](https://revana.si/models/grok-imagine-image-2.0.md): from $0.02 per image

## Grok Imagine Video API

> xAI's Grok Imagine Video: text to video and image to video, up to 1080p, with sound. By xAI, on Revana from $0.02 per second of video.

- Model id: `grok-imagine-video`
- Kind: video
- Maker: xAI
- Tasks: Text to video, Image to video
- Resolutions: 480p, 720p, 1080p
- Durations: 6 to 30 seconds
- Audio: supported (`"audio": true`)
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/grok-imagine-video

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.02 | 2 |
| 720p | $0.03 | 3 |
| 1080p | $0.05 | 5 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p |
|---|---|---|
| Revana | $0.02 | $0.03 |
| xAI's own API | $0.05 | $0.07 |
| fal | $0.05 | $0.07 |
| Replicate | $0.05 | $0.05 |
| CometAPI | $0.04 | $0.056 |
| APIMart | $0.04 | $0.056 |
| WaveSpeed | $0.05 | $0.05 |
| OpenRouter | $0.05 | $0.07 |
| Vercel AI Gateway | $0.05 | $0.07 |
| Runware | $0.05 | $0.07 |

Full comparison and sources: https://revana.si/blog/grok-imagine-video-api-pricing (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "grok-imagine-video",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 6,
    "audio": true
  },
  "max_cost_usd": 0.27
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.18). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Grok Imagine Extend](https://revana.si/models/grok-imagine-extend.md): from $0.02 per second of video
- [Grok Imagine Upscale](https://revana.si/models/grok-imagine-upscale.md): from $0.07 per generation
- [Grok Imagine Video 1.5](https://revana.si/models/grok-imagine-video-1.5.md): from $0.04 per second of video
- [Grok Imagine Video 1.5 Preview](https://revana.si/models/grok-imagine-video-1.5-preview.md): from $0.02 per second of video
- [Grok Imagine Image](https://revana.si/models/grok-imagine-image.md): from $0.02 per image
- [Grok Imagine Image 2.0](https://revana.si/models/grok-imagine-image-2.0.md): from $0.02 per image

## Grok Imagine Video 1.5 API

> xAI's Grok Imagine Video 1.5: text to video, image to video, first and last frames and reference to video, up to 1080p, with sound. By xAI, on Revana from $0.04 per second of video. The lowest listed Grok Imagine Video 1.5 price we found at 480p, 720p and 1080p (checked 2026-10-26).

- Model id: `grok-imagine-video-1.5`
- Kind: video
- Maker: xAI
- Tasks: Text to video, Image to video, First + last frame, Reference to video
- Resolutions: 480p, 720p, 1080p
- Durations: 1 to 15 seconds
- Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, 2:3
- Audio: supported (`"audio": true`)
- Input images: up to 2
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/grok-imagine-video-1.5

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.04 | 4 |
| 720p | $0.07 | 7 |
| 1080p | $0.13 | 13 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-26)

Per second of output for a 16:9 clip with sound. Token prices are converted at width × height × 24 / 1024 tokens a second; card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 480p | 720p | 1080p |
|---|---|---|---|
| Revana | $0.04 | $0.07 | $0.13 |
| xAI's own API | $0.08 | $0.14 | $0.25 |
| fal | $0.08 | $0.14 | $0.25 |
| Replicate | $0.08 | $0.08 | – |
| Runway API | $0.10 | $0.16 | $0.29 |
| CometAPI | $0.064 | $0.112 | $0.20 |
| APIMart | $0.064 | $0.112 | $0.20 |
| WaveSpeed | $0.08 | $0.14 | $0.25 |
| OpenRouter | $0.08 | $0.14 | $0.25 |
| Vercel AI Gateway | $0.08 | $0.14 | $0.25 |
| Runware | $0.08 | $0.14 | $0.25 |

Revana has the lowest listed price per second of every API we checked at 480p ($0.04), 720p ($0.07) and 1080p ($0.13).
Full comparison and sources: https://revana.si/blog/grok-imagine-video-api-pricing (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "grok-imagine-video-1.5",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 1,
    "aspect_ratio": "16:9",
    "audio": true
  },
  "max_cost_usd": 0.11
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.07). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Grok Imagine Extend](https://revana.si/models/grok-imagine-extend.md): from $0.02 per second of video
- [Grok Imagine Upscale](https://revana.si/models/grok-imagine-upscale.md): from $0.07 per generation
- [Grok Imagine Video](https://revana.si/models/grok-imagine-video.md): from $0.02 per second of video
- [Grok Imagine Video 1.5 Preview](https://revana.si/models/grok-imagine-video-1.5-preview.md): from $0.02 per second of video
- [Grok Imagine Image](https://revana.si/models/grok-imagine-image.md): from $0.02 per image
- [Grok Imagine Image 2.0](https://revana.si/models/grok-imagine-image-2.0.md): from $0.02 per image

## Grok Imagine Video 1.5 Preview API

> xAI's Grok Imagine Video 1.5 Preview: text to video and image to video, up to 720p, with sound. By xAI, on Revana from $0.02 per second of video.

- Model id: `grok-imagine-video-1.5-preview`
- Kind: video
- Maker: xAI
- Tasks: Text to video, Image to video
- Resolutions: 480p, 720p
- Durations: 1 to 15 seconds
- Audio: supported (`"audio": true`)
- Input images: up to 1
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/grok-imagine-video-1.5-preview

### Price (per second of video, USD)

| Resolution | Price | Credits |
|---|---|---|
| 480p | $0.02 | 2 |
| 720p | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "grok-imagine-video-1.5-preview",
  "input": {
    "prompt": "A slow dolly shot through a neon-lit night market in the rain, steam rising from food stalls",
    "resolution": "720p",
    "duration": 1,
    "audio": true
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Grok Imagine Extend](https://revana.si/models/grok-imagine-extend.md): from $0.02 per second of video
- [Grok Imagine Upscale](https://revana.si/models/grok-imagine-upscale.md): from $0.07 per generation
- [Grok Imagine Video](https://revana.si/models/grok-imagine-video.md): from $0.02 per second of video
- [Grok Imagine Video 1.5](https://revana.si/models/grok-imagine-video-1.5.md): from $0.04 per second of video
- [Grok Imagine Image](https://revana.si/models/grok-imagine-image.md): from $0.02 per image
- [Grok Imagine Image 2.0](https://revana.si/models/grok-imagine-image-2.0.md): from $0.02 per image

## Qwen Image 2.0 API

> Alibaba's Qwen Image 2.0: text to image and image editing, up to 2K. By Alibaba, on Revana $0.03 per image.

- Model id: `qwen-image-2.0`
- Kind: image
- Maker: Alibaba
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K
- Input images: up to 3
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/qwen-image-2.0

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "qwen-image-2.0",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Qwen Image 2.0 Pro API

> Alibaba's Qwen Image 2.0 Pro: text to image and image editing, up to 2K. By Alibaba, on Revana $0.07 per image.

- Model id: `qwen-image-2.0-pro`
- Kind: image
- Maker: Alibaba
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/qwen-image-2.0-pro

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.07 | 7 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "qwen-image-2.0-pro",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.11
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.07). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Qwen Image 2.1 API

> Alibaba's Qwen Image 2.1: text to image and image editing, up to 2K. By Alibaba, on Revana from $0.03 per image.

- Model id: `qwen-image-2.1`
- Kind: image
- Maker: Alibaba
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K
- Input images: up to 10
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/qwen-image-2.1

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.03 | 3 |
| 2K | $0.05 | 5 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "qwen-image-2.1",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Qwen Image 3.0 API

> Alibaba's Qwen Image 3.0: text to image and image editing, up to 2K. By Alibaba, on Revana from $0.03 per image. The lowest listed Qwen Image 3.0 price we found at 1K and 2K (checked 2026-10-26).

- Model id: `qwen-image-3.0`
- Kind: image
- Maker: Alibaba
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K
- Input images: up to 3
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/qwen-image-3.0

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.03 | 3 |
| 2K | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-30)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K |
|---|---|---|
| Revana | $0.03 | $0.03 |
| Higgsfield API | $0.04 | $0.075 |

Higgsfield prices are its published maximum-discount API rates; eligibility may require selecting the model, and offers can change. Source: https://open.higgsfield.ai/pricing

Revana has the lowest listed price per image of every API we checked at 1K ($0.03) and 2K ($0.03).

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "qwen-image-3.0",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Qwen Image 3.0 Pro API

> Alibaba's Qwen Image 3.0 Pro: text to image and image editing, up to 2K. By Alibaba, on Revana from $0.04 per image.

- Model id: `qwen-image-3.0-pro`
- Kind: image
- Maker: Alibaba
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K
- Input images: up to 3
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/qwen-image-3.0-pro

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.04 | 4 |
| 2K | $0.08 | 8 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "qwen-image-3.0-pro",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.06
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.04). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Qwen Image Edit API

> Alibaba's Qwen Image Edit: image editing. By Alibaba, on Revana $0.04 per image.

- Model id: `qwen-image-edit`
- Kind: image
- Maker: Alibaba
- Tasks: Image to image
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/qwen-image-edit

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.04 | 4 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "qwen-image-edit",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "image_url": "https://example.com/start-frame.jpg"
  },
  "max_cost_usd": 0.06
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.04). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Wan 2.7 Image API

> Alibaba's Wan 2.7 Image: text to image and image editing. By Alibaba, on Revana $0.03 per image.

- Model id: `wan-2.7-image`
- Kind: image
- Maker: Alibaba
- Tasks: Text to image, Image to image
- Input images: up to 9
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/wan-2.7-image

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "wan-2.7-image",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Wan 2.7 Image Pro API

> Alibaba's Wan 2.7 Image Pro: text to image and image editing, up to 4K. By Alibaba, on Revana $0.07 per image.

- Model id: `wan-2.7-image-pro`
- Kind: image
- Maker: Alibaba
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K, 4K
- Input images: up to 9
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/wan-2.7-image-pro

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.07 | 7 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "wan-2.7-image-pro",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.11
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.07). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Z Image API

> Alibaba's Z Image: text to image. By Alibaba, on Revana $0.01 per image.

- Model id: `z-image`
- Kind: image
- Maker: Alibaba
- Tasks: Text to image
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/z-image

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.01 | 1 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "z-image",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light"
  },
  "max_cost_usd": 0.02
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.01). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Z Image Turbo API

> Alibaba's Z Image Turbo: text to image, up to 2K. By Alibaba, on Revana $0.02 per image.

- Model id: `z-image-turbo`
- Kind: image
- Maker: Alibaba
- Tasks: Text to image
- Resolutions: 1K, 2K
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/z-image-turbo

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.02 | 2 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "z-image-turbo",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.03
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.02). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Wan 3.0](https://revana.si/models/wan-3.0.md): from $0.04 per second of video
- [Wan 3.0 Prime](https://revana.si/models/wan-3.0-prime.md): from $0.07 per second of video
- [Wan 2.6](https://revana.si/models/wan-2.6.md): from $0.07 per second of video
- [Qwen Image](https://revana.si/models/qwen-image.md): from $0.03 per image
- [HappyHorse 1.0](https://revana.si/models/happyhorse-1.0.md): from $0.17 per second of video
- [HappyHorse 1.1](https://revana.si/models/happyhorse-1.1.md): from $0.15 per second of video

## Flux 2 Flex API

> Black Forest Labs's Flux 2 Flex: text to image and image editing, up to 4MP. By Black Forest Labs, on Revana from $0.05 per image.

- Model id: `flux-2-flex`
- Kind: image
- Maker: Black Forest Labs
- Tasks: Text to image, Image to image
- Resolutions: 1MP, 2MP, 3MP, 4MP
- Aspect ratios: auto, 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 21:9, 9:21
- Input images: up to 8
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/flux-2-flex

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1MP | $0.05 | 5 |
| 2MP | $0.10 | 10 |
| 3MP | $0.15 | 15 |
| 4MP | $0.20 | 20 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-28)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1MP | 2MP | 3MP | 4MP |
|---|---|---|---|---|
| Revana | $0.05 | $0.10 | $0.15 | $0.20 |
| Replicate | $0.06 | $0.12 | $0.18 | $0.24 |

Full comparison and sources: https://revana.si/blog/cheapest-ai-image-generation-api (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "flux-2-flex",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1MP",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.08
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.05). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Flux 2 Pro](https://revana.si/models/flux-2-pro.md): from $0.03 per image
- [Flux 3 Video](https://revana.si/models/flux-3-video.md): from $0.06 per second of video
- [Flux 2 Max](https://revana.si/models/flux-2-max.md): from $0.07 per image
- [Flux 3](https://revana.si/models/flux-3.md): from $0.05 per image
- [Flux 3 Klein](https://revana.si/models/flux-3-klein.md): from $0.01 per image
- [Flux Kontext Max](https://revana.si/models/flux-kontext-max.md): from $0.07 per image

## Flux 2 Max API

> Black Forest Labs's Flux 2 Max: text to image and image editing, up to 4MP. By Black Forest Labs, on Revana from $0.07 per image.

- Model id: `flux-2-max`
- Kind: image
- Maker: Black Forest Labs
- Tasks: Text to image, Image to image
- Resolutions: 1MP, 2MP, 3MP, 4MP
- Aspect ratios: auto, 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 21:9, 9:21
- Input images: up to 8
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/flux-2-max

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1MP | $0.07 | 7 |
| 2MP | $0.10 | 10 |
| 3MP | $0.13 | 13 |
| 4MP | $0.16 | 16 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "flux-2-max",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1MP",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.11
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.07). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Flux 2 Pro](https://revana.si/models/flux-2-pro.md): from $0.03 per image
- [Flux 3 Video](https://revana.si/models/flux-3-video.md): from $0.06 per second of video
- [Flux 2 Flex](https://revana.si/models/flux-2-flex.md): from $0.05 per image
- [Flux 3](https://revana.si/models/flux-3.md): from $0.05 per image
- [Flux 3 Klein](https://revana.si/models/flux-3-klein.md): from $0.01 per image
- [Flux Kontext Max](https://revana.si/models/flux-kontext-max.md): from $0.07 per image

## Flux 3 API

> Black Forest Labs's Flux 3: text to image and image editing, up to 4K. By Black Forest Labs, on Revana from $0.05 per image.

- Model id: `flux-3`
- Kind: image
- Maker: Black Forest Labs
- Tasks: Text to image, Image to image
- Resolutions: 768SQ, 1K, 1.5K, 2K, 4K
- Input images: up to 10
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/flux-3

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 768SQ | $0.05 | 5 |
| 1K | $0.05 | 5 |
| 1.5K | $0.07 | 7 |
| 2K | $0.10 | 10 |
| 4K | $0.61 | 61 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "flux-3",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.08
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.05). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Flux 2 Pro](https://revana.si/models/flux-2-pro.md): from $0.03 per image
- [Flux 3 Video](https://revana.si/models/flux-3-video.md): from $0.06 per second of video
- [Flux 2 Flex](https://revana.si/models/flux-2-flex.md): from $0.05 per image
- [Flux 2 Max](https://revana.si/models/flux-2-max.md): from $0.07 per image
- [Flux 3 Klein](https://revana.si/models/flux-3-klein.md): from $0.01 per image
- [Flux Kontext Max](https://revana.si/models/flux-kontext-max.md): from $0.07 per image

## Flux 3 Klein API

> Black Forest Labs's Flux 3 Klein: text to image and image editing, up to 4K. By Black Forest Labs, on Revana from $0.01 per image.

- Model id: `flux-3-klein`
- Kind: image
- Maker: Black Forest Labs
- Tasks: Text to image, Image to image
- Resolutions: 768SQ, 1K, 1.5K, 2K, 4K
- Aspect ratios: 1:1, 4:3, 3:4, 3:2, 2:3, 16:9, 9:16, 21:9, 9:21, auto
- Input images: up to 10
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/flux-3-klein

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 768SQ | $0.01 | 1 |
| 1K | $0.01 | 1 |
| 1.5K | $0.02 | 2 |
| 2K | $0.02 | 2 |
| 4K | $0.04 | 4 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "flux-3-klein",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.02
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.01). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Flux 2 Pro](https://revana.si/models/flux-2-pro.md): from $0.03 per image
- [Flux 3 Video](https://revana.si/models/flux-3-video.md): from $0.06 per second of video
- [Flux 2 Flex](https://revana.si/models/flux-2-flex.md): from $0.05 per image
- [Flux 2 Max](https://revana.si/models/flux-2-max.md): from $0.07 per image
- [Flux 3](https://revana.si/models/flux-3.md): from $0.05 per image
- [Flux Kontext Max](https://revana.si/models/flux-kontext-max.md): from $0.07 per image

## Flux Kontext Max API

> Black Forest Labs's Flux Kontext Max: text to image and image editing. By Black Forest Labs, on Revana $0.07 per image.

- Model id: `flux-kontext-max`
- Kind: image
- Maker: Black Forest Labs
- Tasks: Text to image, Image to image
- Input images: up to 4
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/flux-kontext-max

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.07 | 7 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "flux-kontext-max",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light"
  },
  "max_cost_usd": 0.11
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.07). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Flux 2 Pro](https://revana.si/models/flux-2-pro.md): from $0.03 per image
- [Flux 3 Video](https://revana.si/models/flux-3-video.md): from $0.06 per second of video
- [Flux 2 Flex](https://revana.si/models/flux-2-flex.md): from $0.05 per image
- [Flux 2 Max](https://revana.si/models/flux-2-max.md): from $0.07 per image
- [Flux 3](https://revana.si/models/flux-3.md): from $0.05 per image
- [Flux 3 Klein](https://revana.si/models/flux-3-klein.md): from $0.01 per image

## Flux Kontext Pro API

> Black Forest Labs's Flux Kontext Pro: text to image and image editing. By Black Forest Labs, on Revana $0.04 per image.

- Model id: `flux-kontext-pro`
- Kind: image
- Maker: Black Forest Labs
- Tasks: Text to image, Image to image
- Input images: up to 4
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/flux-kontext-pro

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.04 | 4 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "flux-kontext-pro",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light"
  },
  "max_cost_usd": 0.06
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.04). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Flux 2 Pro](https://revana.si/models/flux-2-pro.md): from $0.03 per image
- [Flux 3 Video](https://revana.si/models/flux-3-video.md): from $0.06 per second of video
- [Flux 2 Flex](https://revana.si/models/flux-2-flex.md): from $0.05 per image
- [Flux 2 Max](https://revana.si/models/flux-2-max.md): from $0.07 per image
- [Flux 3](https://revana.si/models/flux-3.md): from $0.05 per image
- [Flux 3 Klein](https://revana.si/models/flux-3-klein.md): from $0.01 per image

## Seedream 3.0 API

> ByteDance's Seedream 3.0: text to image. By ByteDance, on Revana $0.03 per image.

- Model id: `seedream-3.0`
- Kind: image
- Maker: ByteDance
- Tasks: Text to image
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/seedream-3.0

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "seedream-3.0",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image

## Seedream 4.0 API

> ByteDance's Seedream 4.0: text to image and image editing, up to 4K. By ByteDance, on Revana $0.03 per image.

- Model id: `seedream-4.0`
- Kind: image
- Maker: ByteDance
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K, 4K
- Input images: up to 10
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/seedream-4.0

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "seedream-4.0",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image

## Seedream 5.0 Flash API

> ByteDance's Seedream 5.0 Flash: text to image and image editing, up to 2K. By ByteDance, on Revana $0.02 per image.

- Model id: `seedream-5.0-flash`
- Kind: image
- Maker: ByteDance
- Tasks: Text to image, Image to image
- Resolutions: 1K, 1.5K, 2K
- Input images: up to 10
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/seedream-5.0-flash

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.02 | 2 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "seedream-5.0-flash",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.03
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.02). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image

## Seedream 5.0 Lite API

> ByteDance's Seedream 5.0 Lite: text to image and image editing, up to 4K. By ByteDance, on Revana from $0.03 per image. The lowest listed Seedream 5.0 Lite price we found at 2K and 3K (checked 2026-10-26).

- Model id: `seedream-5.0-lite`
- Kind: image
- Maker: ByteDance
- Tasks: Text to image, Image to image
- Resolutions: 2K, 3K, 4K
- Aspect ratios: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 5:4, 4:5, 21:9
- Input images: up to 6
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/seedream-5.0-lite

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 2K | $0.03 | 3 |
| 3K | $0.03 | 3 |
| 4K | $0.04 | 4 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-30)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 2K | 3K |
|---|---|---|
| Revana | $0.03 | $0.03 |
| Replicate | $0.035 | $0.035 |

Revana has the lowest listed price per image of every API we checked at 2K ($0.03) and 3K ($0.03).
Full comparison and sources: https://revana.si/blog/cheapest-ai-image-generation-api (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "seedream-5.0-lite",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "2K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Seedance 2.5](https://revana.si/models/seedance-2.5.md): from $0.09 per second of video
- [Seedance 2.5 Fast](https://revana.si/models/seedance-2.5-fast.md): from $0.10 per second of video
- [Seedance 2.0](https://revana.si/models/seedance-2.0.md): from $0.08 per second of video
- [Seedream 5.5](https://revana.si/models/seedream-5.5.md): from $0.02 per image
- [Seedream 4.5](https://revana.si/models/seedream-4.5.md): from $0.04 per image
- [Seedream 5.0 Pro](https://revana.si/models/seedream-5.0-pro.md): from $0.04 per image

## Imagen 4 API

> Google's Imagen 4: text to image. By Google, on Revana $0.05 per image.

- Model id: `imagen-4`
- Kind: image
- Maker: Google
- Tasks: Text to image
- Aspect ratios: 1:1, 4:3, 3:4, 16:9, 9:16
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/imagen-4

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.05 | 5 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "imagen-4",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.08
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.05). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video
- [Veo 3.1 Fast](https://revana.si/models/veo-3.1-fast.md): from $0.08 per second of video
- [Veo 3.1](https://revana.si/models/veo-3.1.md): from $0.20 per second of video
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image

## Imagen 4 Fast API

> Google's Imagen 4 Fast: text to image. By Google, on Revana $0.03 per image.

- Model id: `imagen-4-fast`
- Kind: image
- Maker: Google
- Tasks: Text to image
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/imagen-4-fast

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "imagen-4-fast",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video
- [Veo 3.1 Fast](https://revana.si/models/veo-3.1-fast.md): from $0.08 per second of video
- [Veo 3.1](https://revana.si/models/veo-3.1.md): from $0.20 per second of video
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image

## Imagen 4 Ultra API

> Google's Imagen 4 Ultra: text to image. By Google, on Revana $0.08 per image.

- Model id: `imagen-4-ultra`
- Kind: image
- Maker: Google
- Tasks: Text to image
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/imagen-4-ultra

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.08 | 8 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "imagen-4-ultra",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light"
  },
  "max_cost_usd": 0.12
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.08). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video
- [Veo 3.1 Fast](https://revana.si/models/veo-3.1-fast.md): from $0.08 per second of video
- [Veo 3.1](https://revana.si/models/veo-3.1.md): from $0.20 per second of video
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image

## Nano Banana 2 Lite API

> Google's Nano Banana 2 Lite: text to image and image editing, up to 1K. By Google, on Revana $0.02 per image.

- Model id: `nano-banana-2-lite`
- Kind: image
- Maker: Google
- Tasks: Text to image, Image to image
- Resolutions: 1K
- Aspect ratios: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 5:4, 4:5, 21:9
- Input images: up to 14
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/nano-banana-2-lite

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.02 | 2 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "nano-banana-2-lite",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.03
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.02). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video
- [Veo 3.1 Fast](https://revana.si/models/veo-3.1-fast.md): from $0.08 per second of video
- [Veo 3.1](https://revana.si/models/veo-3.1.md): from $0.20 per second of video
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image

## Nano Banana 2.1 API

> Google's Nano Banana 2.1: text to image and image editing, up to 4K. By Google, on Revana from $0.03 per image. The lowest listed Nano Banana 2.1 price we found at 4K (checked 2026-10-26).

- Model id: `nano-banana-2.1`
- Kind: image
- Maker: Google
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K, 4K
- Aspect ratios: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 5:4, 4:5, 21:9, 1:4, 4:1, 1:8, 8:1
- Input images: up to 14
- Routes: 3 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/nano-banana-2.1

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.03 | 3 |
| 2K | $0.03 | 3 |
| 4K | $0.04 | 4 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-30)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 4K |
|---|---|
| Revana | $0.04 |
| Kie.ai | $0.045 |

Revana has the lowest listed price per image of every API we checked at 4K ($0.04).

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "nano-banana-2.1",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [Gemini Omni 1.2 Flash](https://revana.si/models/gemini-omni-1.2-flash.md): from $0.024 per second of video
- [Veo 3.1 Fast](https://revana.si/models/veo-3.1-fast.md): from $0.08 per second of video
- [Veo 3.1](https://revana.si/models/veo-3.1.md): from $0.20 per second of video
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image

## Marketing Studio Image API

> Higgsfield's Marketing Studio Image: text to image. By Higgsfield, on Revana from $0.04 per image.

- Model id: `marketing-studio-image`
- Kind: image
- Maker: Higgsfield
- Tasks: Text to image
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/marketing-studio-image

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| high | $0.59 | 59 |
| low | $0.04 | 4 |
| medium | $0.16 | 16 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "marketing-studio-image",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light"
  },
  "max_cost_usd": 0.24
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.16). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image
- [Genjutsu Motion Transfer](https://revana.si/models/genjutsu-motion-transfer.md): from $0.21 per second of video
- [Genjutsu Object Swap](https://revana.si/models/genjutsu-object-swap.md): from $0.21 per second of video
- [Soul](https://revana.si/models/soul.md): from $0.13 per image
- [Soul Cinema](https://revana.si/models/soul-cinema.md): from $0.01 per image
- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image

## Soul API

> Higgsfield's Soul: text to image, up to 1080p. By Higgsfield, on Revana from $0.13 per image.

- Model id: `soul`
- Kind: image
- Maker: Higgsfield
- Tasks: Text to image
- Resolutions: 720p, 1080p
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/soul

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 720p | $0.13 | 13 |
| 1080p | $0.26 | 26 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "soul",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "720p"
  },
  "max_cost_usd": 0.2
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.13). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image
- [Genjutsu Motion Transfer](https://revana.si/models/genjutsu-motion-transfer.md): from $0.21 per second of video
- [Genjutsu Object Swap](https://revana.si/models/genjutsu-object-swap.md): from $0.21 per second of video
- [Marketing Studio Image](https://revana.si/models/marketing-studio-image.md): from $0.04 per image
- [Soul Cinema](https://revana.si/models/soul-cinema.md): from $0.01 per image
- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image

## Soul Cinema API

> Higgsfield's Soul Cinema: film-still pictures with cinematic light and color grading, at 720p or 1080p. By Higgsfield, on Revana from $0.01 per image.

- Model id: `soul-cinema`
- Kind: image
- Maker: Higgsfield
- Tasks: Text to image
- Resolutions: 720p, 1080p
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/soul-cinema

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 720p | $0.01 | 1 |
| 1080p | $0.01 | 1 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "soul-cinema",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "720p"
  },
  "max_cost_usd": 0.02
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.01). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image
- [Genjutsu Motion Transfer](https://revana.si/models/genjutsu-motion-transfer.md): from $0.21 per second of video
- [Genjutsu Object Swap](https://revana.si/models/genjutsu-object-swap.md): from $0.21 per second of video
- [Marketing Studio Image](https://revana.si/models/marketing-studio-image.md): from $0.04 per image
- [Soul](https://revana.si/models/soul.md): from $0.13 per image
- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image

## Ideogram Character API

> Ideogram's Ideogram Character: image editing. By Ideogram, on Revana from $0.08 per image.

- Model id: `ideogram-character`
- Kind: image
- Maker: Ideogram
- Tasks: Image to image
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/ideogram-character

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| balanced | $0.12 | 12 |
| quality | $0.15 | 15 |
| turbo | $0.08 | 8 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "ideogram-character",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "image_url": "https://example.com/start-frame.jpg"
  },
  "max_cost_usd": 0.18
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.12). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Ideogram V3](https://revana.si/models/ideogram-v3.md): from $0.03 per image
- [Ideogram V4](https://revana.si/models/ideogram-v4.md): from $0.04 per image
- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image

## Ideogram V4 API

> Ideogram's Ideogram V4: text to image and image editing. By Ideogram, on Revana from $0.04 per image.

- Model id: `ideogram-v4`
- Kind: image
- Maker: Ideogram
- Tasks: Text to image, Image to image
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/ideogram-v4

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| balanced | $0.08 | 8 |
| quality | $0.14 | 14 |
| turbo | $0.04 | 4 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "ideogram-v4",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light"
  },
  "max_cost_usd": 0.12
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.08). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Ideogram V3](https://revana.si/models/ideogram-v3.md): from $0.03 per image
- [Ideogram Character](https://revana.si/models/ideogram-character.md): from $0.08 per image
- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image

## MAI Image 2.6 API

> Microsoft's MAI Image 2.6: text to image and image editing, up to 2K. By Microsoft, on Revana from $0.04 per image.

- Model id: `mai-image-2.6`
- Kind: image
- Maker: Microsoft
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K
- Input images: up to 5
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/mai-image-2.6

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.04 | 4 |
| 2K | $0.09 | 9 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "mai-image-2.6",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.06
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.04). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [MAI Image 2.6 Flash](https://revana.si/models/mai-image-2.6-flash.md): from $0.02 per image
- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image

## MAI Image 2.6 Flash API

> Microsoft's MAI Image 2.6 Flash: text to image and image editing, up to 2K. By Microsoft, on Revana from $0.02 per image.

- Model id: `mai-image-2.6-flash`
- Kind: image
- Maker: Microsoft
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K
- Input images: up to 5
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/mai-image-2.6-flash

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.02 | 2 |
| 2K | $0.05 | 5 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "mai-image-2.6-flash",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.03
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.02). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [MAI Image 2.6](https://revana.si/models/mai-image-2.6.md): from $0.04 per image
- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image

## MiniMax H3 Image API

> MiniMax's H3 picture model: text to image and edits from up to 4 reference pictures, 1K to 4K, with the prompt rewritten for detail. By MiniMax, on Revana from $0.02 per image.

- Model id: `minimax-h3-image`
- Kind: image
- Maker: MiniMax
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K, 4K
- Input images: up to 4
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/minimax-h3-image

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.02 | 2 |
| 2K | $0.02 | 2 |
| 4K | $0.04 | 4 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "minimax-h3-image",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.03
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.02). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Hailuo 2.3](https://revana.si/models/hailuo-2.3.md): from $0.06 per second of video
- [Hailuo 02](https://revana.si/models/hailuo-02.md): from $0.02 per second of video
- [Hailuo 02 Pro](https://revana.si/models/hailuo-02-pro.md): from $0.06 per second of video
- [Hailuo 02 Standard](https://revana.si/models/hailuo-02-std.md): from $0.02 per second of video
- [Hailuo 2.3 Fast](https://revana.si/models/hailuo-2.3-fast.md): from $0.04 per second of video
- [Hailuo 2.3 Standard](https://revana.si/models/hailuo-2.3-std.md): from $0.04 per second of video

## GPT 4o Image API

> OpenAI's GPT 4o Image: text to image and image editing. By OpenAI, on Revana $0.04 per image.

- Model id: `gpt-4o-image`
- Kind: image
- Maker: OpenAI
- Tasks: Text to image, Image to image
- Input images: up to 5
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/gpt-4o-image

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.04 | 4 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "gpt-4o-image",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light"
  },
  "max_cost_usd": 0.06
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.04). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [GPT Image 1.5](https://revana.si/models/gpt-image-1.5.md): from $0.01 per image
- [GPT Image 1](https://revana.si/models/gpt-image-1.md): from $0.01 per image
- [GPT Image 2.5 Flare](https://revana.si/models/gpt-image-2.5-flare.md): from $0.02 per image
- [GPT Image 2.5 Sunburst](https://revana.si/models/gpt-image-2.5-sunburst.md): from $0.02 per image
- [GPT Image 2.6 Flare](https://revana.si/models/gpt-image-2.6-flare.md): from $0.02 per image

## GPT Image 1 API

> OpenAI's GPT Image 1: text to image and image editing. By OpenAI, on Revana from $0.01 per image.

- Model id: `gpt-image-1`
- Kind: image
- Maker: OpenAI
- Tasks: Text to image, Image to image
- Aspect ratios: 1:1, 3:2, 2:3
- Input images: up to 15
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/gpt-image-1

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1:1-high | $0.14 | 14 |
| 1:1-low | $0.01 | 1 |
| 1:1-medium | $0.04 | 4 |
| 2:3-high | $0.20 | 20 |
| 2:3-low | $0.02 | 2 |
| 2:3-medium | $0.06 | 6 |
| 3:2-high | $0.20 | 20 |
| 3:2-low | $0.02 | 2 |
| 3:2-medium | $0.06 | 6 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "gpt-image-1",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "aspect_ratio": "1:1"
  },
  "max_cost_usd": 0.06
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.04). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [GPT Image 1.5](https://revana.si/models/gpt-image-1.5.md): from $0.01 per image
- [GPT 4o Image](https://revana.si/models/gpt-4o-image.md): from $0.04 per image
- [GPT Image 2.5 Flare](https://revana.si/models/gpt-image-2.5-flare.md): from $0.02 per image
- [GPT Image 2.5 Sunburst](https://revana.si/models/gpt-image-2.5-sunburst.md): from $0.02 per image
- [GPT Image 2.6 Flare](https://revana.si/models/gpt-image-2.6-flare.md): from $0.02 per image

## GPT Image 2.5 Flare API

> OpenAI's GPT Image 2.5 Flare: text to image and image editing, up to 4K. By OpenAI, on Revana from $0.02 per image. The lowest listed GPT Image 2.5 Flare price we found at 1K, 2K and 4K (checked 2026-10-26).

- Model id: `gpt-image-2.5-flare`
- Kind: image
- Maker: OpenAI
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K, 4K
- Aspect ratios: 1:1, 3:4, 4:3, 2:3, 3:2, 9:16, 16:9, 21:9, 5:2
- Input images: up to 16
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/gpt-image-2.5-flare

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.02 | 2 |
| 2K | $0.02 | 2 |
| 4K | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-30)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K | 4K |
|---|---|---|---|
| Revana | $0.02 | $0.02 | $0.03 |
| Kie.ai | $0.03 | $0.05 | $0.08 |

Revana has the lowest listed price per image of every API we checked at 1K ($0.02), 2K ($0.02) and 4K ($0.03).

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "gpt-image-2.5-flare",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.03
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.02). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [GPT Image 1.5](https://revana.si/models/gpt-image-1.5.md): from $0.01 per image
- [GPT 4o Image](https://revana.si/models/gpt-4o-image.md): from $0.04 per image
- [GPT Image 1](https://revana.si/models/gpt-image-1.md): from $0.01 per image
- [GPT Image 2.5 Sunburst](https://revana.si/models/gpt-image-2.5-sunburst.md): from $0.02 per image
- [GPT Image 2.6 Flare](https://revana.si/models/gpt-image-2.6-flare.md): from $0.02 per image

## GPT Image 2.5 Sunburst API

> OpenAI's GPT Image 2.5 Sunburst: text to image and image editing, up to 4K. By OpenAI, on Revana from $0.02 per image. The lowest listed GPT Image 2.5 Sunburst price we found at 1K, 2K and 4K (checked 2026-10-26).

- Model id: `gpt-image-2.5-sunburst`
- Kind: image
- Maker: OpenAI
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K, 4K
- Aspect ratios: 1:1, 3:4, 4:3, 2:3, 3:2, 9:16, 16:9, 21:9, 5:2
- Input images: up to 16
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/gpt-image-2.5-sunburst

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.02 | 2 |
| 2K | $0.02 | 2 |
| 4K | $0.03 | 3 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-30)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K | 4K |
|---|---|---|---|
| Revana | $0.02 | $0.02 | $0.03 |
| Kie.ai | $0.03 | $0.05 | $0.08 |

Revana has the lowest listed price per image of every API we checked at 1K ($0.02), 2K ($0.02) and 4K ($0.03).

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "gpt-image-2.5-sunburst",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.03
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.02). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [GPT Image 1.5](https://revana.si/models/gpt-image-1.5.md): from $0.01 per image
- [GPT 4o Image](https://revana.si/models/gpt-4o-image.md): from $0.04 per image
- [GPT Image 1](https://revana.si/models/gpt-image-1.md): from $0.01 per image
- [GPT Image 2.5 Flare](https://revana.si/models/gpt-image-2.5-flare.md): from $0.02 per image
- [GPT Image 2.6 Flare](https://revana.si/models/gpt-image-2.6-flare.md): from $0.02 per image

## GPT Image 2.6 Flare API

> OpenAI's GPT Image 2.6 Flare: text to image and image editing, up to 2K. By OpenAI, on Revana from $0.02 per image. The lowest listed GPT Image 2.6 Flare price we found at 1K and 2K (checked 2026-10-26).

- Model id: `gpt-image-2.6-flare`
- Kind: image
- Maker: OpenAI
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K
- Aspect ratios: 1:1, 3:4, 4:3, 2:3, 3:2, 9:16, 16:9, 21:9, 5:2
- Input images: up to 16
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/gpt-image-2.6-flare

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.02 | 2 |
| 2K | $0.02 | 2 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-11-05)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K |
|---|---|---|
| Revana | $0.02 | $0.02 |
| Kie.ai | $0.03 | $0.05 |

Revana has the lowest listed price per image of every API we checked at 1K ($0.02) and 2K ($0.02).
Full comparison and sources: https://revana.si/blog/gpt-image-2-6-api (data: https://revana.si/data/market-prices.json)

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "gpt-image-2.6-flare",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.03
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.02). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [GPT Image 1.5](https://revana.si/models/gpt-image-1.5.md): from $0.01 per image
- [GPT 4o Image](https://revana.si/models/gpt-4o-image.md): from $0.04 per image
- [GPT Image 1](https://revana.si/models/gpt-image-1.md): from $0.01 per image
- [GPT Image 2.5 Flare](https://revana.si/models/gpt-image-2.5-flare.md): from $0.02 per image
- [GPT Image 2.5 Sunburst](https://revana.si/models/gpt-image-2.5-sunburst.md): from $0.02 per image

## GPT Image 2.6 Sunburst API

> OpenAI's GPT Image 2.6 Sunburst: text to image and image editing, up to 2K. By OpenAI, on Revana from $0.02 per image. The lowest listed GPT Image 2.6 Sunburst price we found at 1K and 2K (checked 2026-10-26).

- Model id: `gpt-image-2.6-sunburst`
- Kind: image
- Maker: OpenAI
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K
- Aspect ratios: 1:1, 3:4, 4:3, 2:3, 3:2, 9:16, 16:9, 21:9, 5:2
- Input images: up to 16
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/gpt-image-2.6-sunburst

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.02 | 2 |
| 2K | $0.02 | 2 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-11-05)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K |
|---|---|---|
| Revana | $0.02 | $0.02 |
| Kie.ai | $0.03 | $0.05 |

Revana has the lowest listed price per image of every API we checked at 1K ($0.02) and 2K ($0.02).

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "gpt-image-2.6-sunburst",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.03
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.02). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [GPT Image 1.5](https://revana.si/models/gpt-image-1.5.md): from $0.01 per image
- [GPT 4o Image](https://revana.si/models/gpt-4o-image.md): from $0.04 per image
- [GPT Image 1](https://revana.si/models/gpt-image-1.md): from $0.01 per image
- [GPT Image 2.5 Flare](https://revana.si/models/gpt-image-2.5-flare.md): from $0.02 per image
- [GPT Image 2.5 Sunburst](https://revana.si/models/gpt-image-2.5-sunburst.md): from $0.02 per image

## Recraft Crisp Upscale API

> Recraft's Recraft Crisp Upscale: upscaling. By Recraft, on Revana $0.01 per image.

- Model id: `recraft-crisp-upscale`
- Kind: image
- Maker: Recraft
- Tasks: Upscale
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/recraft-crisp-upscale

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.01 | 1 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "recraft-crisp-upscale",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "image_url": "https://example.com/start-frame.jpg"
  },
  "max_cost_usd": 0.02
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.01). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Recraft V4.1](https://revana.si/models/recraft-v4.1.md): from $0.05 per image
- [Recraft V4.1 Utility](https://revana.si/models/recraft-v4.1-utility.md): from $0.05 per image
- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image

## Recraft V4.1 API

> Recraft's Recraft V4.1: text to image, up to 2K. By Recraft, on Revana from $0.05 per image.

- Model id: `recraft-v4.1`
- Kind: image
- Maker: Recraft
- Tasks: Text to image
- Resolutions: 1K, 2K
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/recraft-v4.1

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.05 | 5 |
| 2K | $0.28 | 28 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "recraft-v4.1",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.08
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.05). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Recraft Crisp Upscale](https://revana.si/models/recraft-crisp-upscale.md): from $0.01 per image
- [Recraft V4.1 Utility](https://revana.si/models/recraft-v4.1-utility.md): from $0.05 per image
- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image

## Recraft V4.1 Utility API

> Recraft's Recraft V4.1 Utility: text to image, up to 2K. By Recraft, on Revana from $0.05 per image.

- Model id: `recraft-v4.1-utility`
- Kind: image
- Maker: Recraft
- Tasks: Text to image
- Resolutions: 1K, 2K
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/recraft-v4.1-utility

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.05 | 5 |
| 2K | $0.28 | 28 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "recraft-v4.1-utility",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.08
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.05). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Recraft Crisp Upscale](https://revana.si/models/recraft-crisp-upscale.md): from $0.01 per image
- [Recraft V4.1](https://revana.si/models/recraft-v4.1.md): from $0.05 per image
- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image

## Hunyuan Image 3.5 Preview API

> Tencent's Hunyuan Image 3.5 Preview: text to image and image editing, up to 4K. By Tencent, on Revana from $0.03 per image.

- Model id: `hunyuan-image-3.5-preview`
- Kind: image
- Maker: Tencent
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K, 4K
- Aspect ratios: auto, 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9
- Input images: up to 5
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/hunyuan-image-3.5-preview

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.03 | 3 |
| 2K | $0.03 | 3 |
| 4K | $0.04 | 4 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "hunyuan-image-3.5-preview",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.05
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.03). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Nano Banana Pro](https://revana.si/models/nano-banana-pro.md): from $0.04 per image
- [Nano Banana 2](https://revana.si/models/nano-banana-2.md): from $0.03 per image
- [GPT Image 2](https://revana.si/models/gpt-image-2.md): from $0.03 per image
- [Soul 2](https://revana.si/models/soul-2.md): from $0.01 per image
- [Nano Banana](https://revana.si/models/nano-banana.md): from $0.03 per image
- [GPT Image 1.5](https://revana.si/models/gpt-image-1.5.md): from $0.01 per image

## Grok Imagine Image API

> xAI's Grok Imagine Image: text to image and image editing, up to 2K. By xAI, on Revana from $0.02 per image.

- Model id: `grok-imagine-image`
- Kind: image
- Maker: xAI
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K
- Aspect ratios: 2:3, 3:2, 1:1, 16:9, 9:16
- Input images: up to 5
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/grok-imagine-image

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.02 | 2 |
| 2K | $0.02 | 2 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "grok-imagine-image",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.03
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.02). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Grok Imagine Extend](https://revana.si/models/grok-imagine-extend.md): from $0.02 per second of video
- [Grok Imagine Upscale](https://revana.si/models/grok-imagine-upscale.md): from $0.07 per generation
- [Grok Imagine Video](https://revana.si/models/grok-imagine-video.md): from $0.02 per second of video
- [Grok Imagine Video 1.5](https://revana.si/models/grok-imagine-video-1.5.md): from $0.04 per second of video
- [Grok Imagine Video 1.5 Preview](https://revana.si/models/grok-imagine-video-1.5-preview.md): from $0.02 per second of video
- [Grok Imagine Image 2.0](https://revana.si/models/grok-imagine-image-2.0.md): from $0.02 per image

## Grok Imagine Image 2.0 API

> xAI's Grok Imagine Image 2.0: text to image and image editing, up to 2K. By xAI, on Revana from $0.02 per image. The lowest listed Grok Imagine Image 2.0 price we found at 1K and 2K (checked 2026-10-26).

- Model id: `grok-imagine-image-2.0`
- Kind: image
- Maker: xAI
- Tasks: Text to image, Image to image
- Resolutions: 1K, 1.5K, 2K
- Aspect ratios: auto, 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 2:1, 1:2, 19.5:9, 9:19.5, 20:9, 9:20, 21:9, 5:2
- Input images: up to 3
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/grok-imagine-image-2.0

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.02 | 2 |
| 1.5K | $0.02 | 2 |
| 2K | $0.02 | 2 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Price at other APIs (checked 2026-10-30)

Per output image at each resolution. Input tokens and reference images, where a seller bills them, and card and top-up fees are not included. Only prices above Revana's are shown; a dash means the API does not list that setting above Revana's price.

| API | 1K | 2K |
|---|---|---|
| Revana | $0.02 | $0.02 |
| Higgsfield API | $0.04 | $0.08 |

Higgsfield prices are its published maximum-discount API rates; eligibility may require selecting the model, and offers can change. Source: https://open.higgsfield.ai/pricing

Revana has the lowest listed price per image of every API we checked at 1K ($0.02) and 2K ($0.02).

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "grok-imagine-image-2.0",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  },
  "max_cost_usd": 0.03
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.02). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Grok Imagine Extend](https://revana.si/models/grok-imagine-extend.md): from $0.02 per second of video
- [Grok Imagine Upscale](https://revana.si/models/grok-imagine-upscale.md): from $0.07 per generation
- [Grok Imagine Video](https://revana.si/models/grok-imagine-video.md): from $0.02 per second of video
- [Grok Imagine Video 1.5](https://revana.si/models/grok-imagine-video-1.5.md): from $0.04 per second of video
- [Grok Imagine Video 1.5 Preview](https://revana.si/models/grok-imagine-video-1.5-preview.md): from $0.02 per second of video
- [Grok Imagine Image](https://revana.si/models/grok-imagine-image.md): from $0.02 per image

## Grok Imagine Image Quality API

> xAI's Grok Imagine Image Quality: text to image and image editing, up to 2K. By xAI, on Revana from $0.05 per image.

- Model id: `grok-imagine-image-quality`
- Kind: image
- Maker: xAI
- Tasks: Text to image, Image to image
- Resolutions: 1K, 2K
- Input images: up to 3
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/grok-imagine-image-quality

### Price (per image, USD)

| Setting | Price | Credits |
|---|---|---|
| 1K | $0.05 | 5 |
| 2K | $0.07 | 7 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "grok-imagine-image-quality",
  "input": {
    "prompt": "A product photo of a ceramic coffee cup on a linen tablecloth, soft morning light",
    "resolution": "1K"
  },
  "max_cost_usd": 0.08
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.05). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [Grok Imagine Extend](https://revana.si/models/grok-imagine-extend.md): from $0.02 per second of video
- [Grok Imagine Upscale](https://revana.si/models/grok-imagine-upscale.md): from $0.07 per generation
- [Grok Imagine Video](https://revana.si/models/grok-imagine-video.md): from $0.02 per second of video
- [Grok Imagine Video 1.5](https://revana.si/models/grok-imagine-video-1.5.md): from $0.04 per second of video
- [Grok Imagine Video 1.5 Preview](https://revana.si/models/grok-imagine-video-1.5-preview.md): from $0.02 per second of video
- [Grok Imagine Image](https://revana.si/models/grok-imagine-image.md): from $0.02 per image

## ElevenLabs Multilingual V2 API

> ElevenLabs' steady, natural voice model for voiceovers and narration in 29 languages, with dozens of voices to pick from. Priced by the characters read. By ElevenLabs, on Revana $0.12 per 1,000 characters.

- Model id: `elevenlabs-multilingual-v2`
- Kind: audio
- Maker: ElevenLabs
- Tasks: Text to speech
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/elevenlabs-multilingual-v2

### Price (per 1,000 characters, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.12 | 12 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "elevenlabs-multilingual-v2",
  "input": {
    "text": "Nobody tells you this, but the hardest part of starting is the first five minutes.",
    "voice": "EkK5I93UQWFDigLMpZcX"
  },
  "max_cost_usd": 0.02
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.01). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [ElevenLabs Turbo 2.5](https://revana.si/models/elevenlabs-turbo-2.5.md): from $0.06 per 1,000 characters
- [ElevenLabs V3](https://revana.si/models/elevenlabs-v3.md): from $0.12 per 1,000 characters
- [ElevenLabs V4](https://revana.si/models/elevenlabs-v4.md): from $0.12 per 1,000 characters
- [Suno V6](https://revana.si/models/suno-v6.md): from $0.08 per generation

## ElevenLabs Turbo 2.5 API

> ElevenLabs' fast, low-cost voice model: quick voiceovers in 32 languages at half the price of Multilingual v2, with language enforcement. Priced by the characters read. By ElevenLabs, on Revana $0.06 per 1,000 characters.

- Model id: `elevenlabs-turbo-2.5`
- Kind: audio
- Maker: ElevenLabs
- Tasks: Text to speech
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/elevenlabs-turbo-2.5

### Price (per 1,000 characters, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.06 | 6 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "elevenlabs-turbo-2.5",
  "input": {
    "text": "Nobody tells you this, but the hardest part of starting is the first five minutes.",
    "voice": "EkK5I93UQWFDigLMpZcX"
  },
  "max_cost_usd": 0.02
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.01). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [ElevenLabs Multilingual V2](https://revana.si/models/elevenlabs-multilingual-v2.md): from $0.12 per 1,000 characters
- [ElevenLabs V3](https://revana.si/models/elevenlabs-v3.md): from $0.12 per 1,000 characters
- [ElevenLabs V4](https://revana.si/models/elevenlabs-v4.md): from $0.12 per 1,000 characters
- [Suno V6](https://revana.si/models/suno-v6.md): from $0.08 per generation

## ElevenLabs V3 API

> ElevenLabs' most expressive voice model: lifelike narration with emotion and audio tags such as [whispers] or [laughs], and dialogue, several voices in one take. Priced by the characters read. By ElevenLabs, on Revana $0.12 per 1,000 characters.

- Model id: `elevenlabs-v3`
- Kind: audio
- Maker: ElevenLabs
- Tasks: Text to speech
- Routes: 2 (starts on the cheapest, falls back automatically)
- Status: live on the API
- Page: https://revana.si/models/elevenlabs-v3

### Price (per 1,000 characters, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.12 | 12 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "elevenlabs-v3",
  "input": {
    "text": "Nobody tells you this, but the hardest part of starting is the first five minutes.",
    "voice": "EkK5I93UQWFDigLMpZcX"
  },
  "max_cost_usd": 0.02
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.01). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [ElevenLabs Multilingual V2](https://revana.si/models/elevenlabs-multilingual-v2.md): from $0.12 per 1,000 characters
- [ElevenLabs Turbo 2.5](https://revana.si/models/elevenlabs-turbo-2.5.md): from $0.06 per 1,000 characters
- [ElevenLabs V4](https://revana.si/models/elevenlabs-v4.md): from $0.12 per 1,000 characters
- [Suno V6](https://revana.si/models/suno-v6.md): from $0.08 per generation

## ElevenLabs V4 API

> ElevenLabs' newest voice model: natural, expressive narration with the steadiness of Multilingual v2, for voiceovers and long reads. Priced by the characters read. By ElevenLabs, on Revana $0.12 per 1,000 characters.

- Model id: `elevenlabs-v4`
- Kind: audio
- Maker: ElevenLabs
- Tasks: Text to speech
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/elevenlabs-v4

### Price (per 1,000 characters, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.12 | 12 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "elevenlabs-v4",
  "input": {
    "text": "Nobody tells you this, but the hardest part of starting is the first five minutes.",
    "voice": "EkK5I93UQWFDigLMpZcX"
  },
  "max_cost_usd": 0.02
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.01). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [ElevenLabs Multilingual V2](https://revana.si/models/elevenlabs-multilingual-v2.md): from $0.12 per 1,000 characters
- [ElevenLabs Turbo 2.5](https://revana.si/models/elevenlabs-turbo-2.5.md): from $0.06 per 1,000 characters
- [ElevenLabs V3](https://revana.si/models/elevenlabs-v3.md): from $0.12 per 1,000 characters
- [Suno V6](https://revana.si/models/suno-v6.md): from $0.08 per generation

## Suno V6 API

> Suno's newest music model: a full song with vocals, or an instrumental, from an idea, your lyrics or a style. Each generation returns two tracks. By Suno, on Revana $0.08 per generation.

- Model id: `suno-v6`
- Kind: audio
- Maker: Suno
- Tasks: Text to music
- Routes: 1
- Status: live on the API
- Page: https://revana.si/models/suno-v6

### Price (per generation, USD)

| Setting | Price | Credits |
|---|---|---|
| All settings | $0.08 | 8 |

100 credits = $1. Charged once, only for the route that succeeds; failed generations cost nothing.

### Request

POST https://api.revana.si/v1/generations with `Authorization: Bearer $REVANA_API_KEY` and an `Idempotency-Key` header: one key per request, made once and sent again on every resend after a timeout, a 429 or a 5xx, so the request is never made or charged twice. After a generation has failed, use a new key.

```json
{
  "model": "suno-v6",
  "input": {
    "prompt": "An upbeat acoustic song about a summer road trip",
    "style": "indie folk, warm vocals",
    "instrumental": false
  },
  "max_cost_usd": 0.12
}
```

`max_cost_usd` is your cap, the most this request may cost (it lists at $0.08). Poll `GET /v1/generations/{id}` until `status` is `succeeded` or `failed`, then download `GET /v1/generations/{id}/content`. Full API: https://revana.si/llms.txt

### Similar models

- [ElevenLabs Multilingual V2](https://revana.si/models/elevenlabs-multilingual-v2.md): from $0.12 per 1,000 characters
- [ElevenLabs Turbo 2.5](https://revana.si/models/elevenlabs-turbo-2.5.md): from $0.06 per 1,000 characters
- [ElevenLabs V3](https://revana.si/models/elevenlabs-v3.md): from $0.12 per 1,000 characters
- [ElevenLabs V4](https://revana.si/models/elevenlabs-v4.md): from $0.12 per 1,000 characters


# Comparisons

## Revana vs Kie.ai

Revana's list price is below Kie.ai's for 34 settings across 13 models: Seedance 2.5, Wan 3.0, Nano Banana Pro, Nano Banana 2, GPT Image 2, Seedance 2.0, Seedream 5.0 Pro, Seedance 2.0 Fast, Nano Banana 2.1, GPT Image 2.5 Flare, GPT Image 2.5 Sunburst, GPT Image 2.6 Flare and GPT Image 2.6 Sunburst (checked 2026-10-30). Kie.ai lists the same price or less for some settings of 24 models, including Wan 3.0 Prime, Veo 3.1 Fast, Veo 3.1 and Kling 3.0.

- Seedance 2.5: $0.09 a second at 480p on Revana, against $0.14 on Kie.ai (36% less).
- Wan 3.0: $0.07 a second at 720p on Revana, against $0.08 on Kie.ai (12% less).
- Nano Banana Pro: $0.04 an image at 1K on Revana, against $0.09 on Kie.ai (56% less).
- Nano Banana 2: $0.04 an image at 4K on Revana, against $0.09 on Kie.ai (56% less).
- GPT Image 2: $0.03 an image at 4K on Revana, against $0.08 on Kie.ai (63% less).

### Popular models at a glance

- Seedance 2.5: Revana is lower at 480p ($0.09 against $0.14, 36% less), at 720p ($0.21 against $0.315, 33% less) and at 1080p ($0.52 against $0.79, 34% less), and at 3 more settings in the tables below.
- Veo 3.1: Kie.ai is cheaper.
- Veo 3.1 Fast: Kie.ai is cheaper.
- Kling 3.0: Kie.ai is cheaper.
- Wan 3.0: Revana is lower at 720p ($0.07 against $0.08, 12% less) and at 1080p ($0.14 against $0.16, 12% less). Kie.ai is cheaper at 480p.
- Nano Banana Pro: Revana is lower at 1K ($0.04 against $0.09, 56% less), at 2K ($0.04 against $0.09, 56% less) and at 4K ($0.08 against $0.12, 33% less).

### Where Kie.ai is cheaper

Kie.ai is cheaper than Revana on 24 models: Veo 3.1, Veo 3.1 Fast, Kling 3.0, Wan 3.0, Wan 3.0 Prime, GPT Image 2, Wan 2.6, Hailuo 2.3, Nano Banana, Seedream 5.0 Pro, Ideogram V3, HappyHorse 1.1, Wan 2.2 Animate, Seedance 2.0 Mini, Veo 3.1 Lite, Kling 2.6 Motion Control, Kling 3.0 Motion Control, Kling 3.0 Turbo, MiniMax H3, PixVerse V6, Grok Imagine Video, Grok Imagine Video 1.5 Preview, Qwen Image 2.1 and Nano Banana 2.1, for at least one setting. This counts Kie.ai's top-up bonus of up to 10%.

### Video models per second (checked 2026-10-30, 2026-11-10)

| Model | Setting | Revana | Kie.ai | Revana is | Source |
|---|---|---|---|---|---|
| Seedance 2.5 | 480p | $0.09 | $0.14 | 36% less | https://kie.ai/seedance-2-5 |
| Seedance 2.5 | 720p | $0.21 | $0.315 | 33% less | https://kie.ai/seedance-2-5 |
| Seedance 2.5 | 1080p | $0.52 | $0.79 | 34% less | https://kie.ai/seedance-2-5 |
| Seedance 2.5 | 480p with a video input | $0.06 | $0.085 | 29% less | https://kie.ai/seedance-2-5 |
| Seedance 2.5 | 720p with a video input | $0.15 | $0.19 | 21% less | https://kie.ai/seedance-2-5 |
| Seedance 2.5 | 1080p with a video input | $0.34 | $0.475 | 28% less | https://kie.ai/seedance-2-5 |
| Wan 3.0 | 720p | $0.07 | $0.08 | 12% less | https://kie.ai/wan3.0-video |
| Wan 3.0 | 1080p | $0.14 | $0.16 | 12% less | https://kie.ai/wan3.0-video |
| Seedance 2.0 | 480p | $0.08 | $0.095 | 16% less | https://kie.ai/seedance-2-0 |
| Seedance 2.0 | 720p | $0.15 | $0.205 | 27% less | https://kie.ai/seedance-2-0 |
| Seedance 2.0 | 1080p | $0.40 | $0.51 | 22% less | https://kie.ai/seedance-2-0 |
| Seedance 2.0 | 4K | $0.83 | $1.04 | 20% less | https://kie.ai/seedance-2-0 |
| Seedance 2.0 Fast | 480p | $0.05 | $0.059 | 15% less | https://kie.ai/seedance-2-0 |
| Seedance 2.0 Fast | 720p | $0.11 | $0.124 | 11% less | https://kie.ai/seedance-2-0 |

### Image models per image (checked 2026-10-30, 2026-11-05)

| Model | Setting | Revana | Kie.ai | Revana is | Source |
|---|---|---|---|---|---|
| Nano Banana Pro | 1K | $0.04 | $0.09 | 56% less | https://kie.ai/nano-banana-pro |
| Nano Banana Pro | 2K | $0.04 | $0.09 | 56% less | https://kie.ai/nano-banana-pro |
| Nano Banana Pro | 4K | $0.08 | $0.12 | 33% less | https://kie.ai/nano-banana-pro |
| Nano Banana 2 | 1K | $0.03 | $0.04 | 25% less | https://kie.ai/nano-banana-2 |
| Nano Banana 2 | 2K | $0.03 | $0.06 | 50% less | https://kie.ai/nano-banana-2 |
| Nano Banana 2 | 4K | $0.04 | $0.09 | 56% less | https://kie.ai/nano-banana-2 |
| GPT Image 2 | 2K | $0.03 | $0.05 | 40% less | https://kie.ai/gpt-image-2 |
| GPT Image 2 | 4K | $0.03 | $0.08 | 63% less | https://kie.ai/gpt-image-2 |
| Seedream 5.0 Pro | 2K | $0.04 | $0.07 | 43% less | https://kie.ai/seedream-5-0-pro |
| Nano Banana 2.1 | 4K | $0.04 | $0.045 | 11% less | https://kie.ai/nano-banana-2-1 |
| GPT Image 2.5 Flare | 1K | $0.02 | $0.03 | 33% less | https://kie.ai/gpt-image-2-5 |
| GPT Image 2.5 Flare | 2K | $0.02 | $0.05 | 60% less | https://kie.ai/gpt-image-2-5 |
| GPT Image 2.5 Flare | 4K | $0.03 | $0.08 | 63% less | https://kie.ai/gpt-image-2-5 |
| GPT Image 2.5 Sunburst | 1K | $0.02 | $0.03 | 33% less | https://kie.ai/gpt-image-2-5 |
| GPT Image 2.5 Sunburst | 2K | $0.02 | $0.05 | 60% less | https://kie.ai/gpt-image-2-5 |
| GPT Image 2.5 Sunburst | 4K | $0.03 | $0.08 | 63% less | https://kie.ai/gpt-image-2-5 |
| GPT Image 2.6 Flare | 1K | $0.02 | $0.03 | 33% less | https://kie.ai/gpt-image-2-6 |
| GPT Image 2.6 Flare | 2K | $0.02 | $0.05 | 60% less | https://kie.ai/gpt-image-2-6 |
| GPT Image 2.6 Sunburst | 1K | $0.02 | $0.03 | 33% less | https://kie.ai/gpt-image-2-6 |
| GPT Image 2.6 Sunburst | 2K | $0.02 | $0.05 | 60% less | https://kie.ai/gpt-image-2-6 |

Kie.ai adds up to 10% bonus credits to larger top-ups, so each Kie.ai price was cut by 10% before comparing: every row here is lower on Revana even with that bonus.

These are Revana's list prices, charged on the first route. For Seedance 2.5, Wan 3.0, Nano Banana Pro, Nano Banana 2, GPT Image 2, Seedream 5.0 Pro, Seedance 2.0 Fast, Nano Banana 2.1, GPT Image 2.5 Flare and GPT Image 2.5 Sunburst, a backup route can cost more if the first one fails; each model page lists every route's price, and max_cost_usd sets the most a job may cost.

### How Revana and Kie.ai differ

Kie.ai resells many video, image and chat models through one API key. It prices them in credits at $0.005 each and adds bonus credits to larger top-ups.

Revana sells 139 video, image and music models and 99 language models through one API key and one prepaid balance. The language models are behind an OpenAI-compatible chat completions endpoint.

Revana routes each model across several providers where it can: it tries the lowest-cost route for the model and setting first and, if that route fails, moves the job to the next one. A job that fails on every route costs nothing, and a prompt the model refuses is not charged.

Credit is prepaid in packs from $10, with no subscription, seats or monthly minimum. Every finished job reports cost_usd, the exact amount charged.

An MCP server lets agents in Claude, ChatGPT, Cursor and Codex price and run generations directly.

### Use it from Claude, ChatGPT or Cursor

Revana has an MCP server, so Claude, ChatGPT, Cursor, VS Code and Codex can use it directly: add https://revana.si/api/mcp as a connector, approve it on revana.si, and ask for a video or an image. The agent lists the models, gets a free quote, tells you the price, makes it on your prepaid balance once you say yes, and hands you the download link.

### Questions

#### Is Revana cheaper than Kie.ai?

For Seedance 2.5, Wan 3.0, Nano Banana Pro, Nano Banana 2, GPT Image 2, Seedance 2.0, Seedream 5.0 Pro, Seedance 2.0 Fast, Nano Banana 2.1, GPT Image 2.5 Flare, GPT Image 2.5 Sunburst, GPT Image 2.6 Flare and GPT Image 2.6 Sunburst, yes: Seedance 2.5: $0.09 a second at 480p on Revana, against $0.14 on Kie.ai (36% less), checked 2026-11-10. Not for every model: Kie.ai lists the same price or less for some settings of 24 models, including Wan 3.0 Prime, Veo 3.1 Fast, Veo 3.1 and Kling 3.0.

#### What is a cheaper Kie.ai alternative for Seedance 2.5, Wan 3.0 and Nano Banana Pro?

Revana. Seedance 2.5: $0.09 a second at 480p on Revana, against $0.14 on Kie.ai (36% less). Wan 3.0: $0.07 a second at 720p on Revana, against $0.08 on Kie.ai (12% less). Nano Banana Pro: $0.04 an image at 1K on Revana, against $0.09 on Kie.ai (56% less). Each price was checked on 2026-10-30 or later.

#### Am I charged for failed generations on Revana?

No. Revana holds the price when you submit and releases it if the job fails, including when the model's safety filter refuses the prompt. If a route fails, the router moves the job to the next route for that model; if none succeeds, the job costs nothing.

#### What happens on Revana when a provider is down?

Revana sends each job to the lowest-cost route for that model and setting and, if that route fails, moves it to the next one, so an outage at one provider does not stop the job where the model has another route. Of the models above, Seedance 2.5, Wan 3.0, Nano Banana Pro, Nano Banana 2, GPT Image 2, Seedance 2.0, Seedream 5.0 Pro, Seedance 2.0 Fast and Nano Banana 2.1 have more than one route.

#### Where is Kie.ai cheaper than Revana?

Kie.ai is cheaper than Revana on 24 models: Veo 3.1, Veo 3.1 Fast, Kling 3.0, Wan 3.0, Wan 3.0 Prime, GPT Image 2, Wan 2.6, Hailuo 2.3, Nano Banana, Seedream 5.0 Pro, Ideogram V3, HappyHorse 1.1, Wan 2.2 Animate, Seedance 2.0 Mini, Veo 3.1 Lite, Kling 2.6 Motion Control, Kling 3.0 Motion Control, Kling 3.0 Turbo, MiniMax H3, PixVerse V6, Grok Imagine Video, Grok Imagine Video 1.5 Preview, Qwen Image 2.1 and Nano Banana 2.1, for at least one setting. This counts Kie.ai's top-up bonus of up to 10%. Every model is listed with its source in the raw price file, so you can check both price lists before moving.

#### Can I use Revana from Claude or ChatGPT?

Revana has an MCP server, so Claude, ChatGPT, Cursor, VS Code and Codex can use it directly: add https://revana.si/api/mcp as a connector, approve it on revana.si, and ask for a video or an image. The agent lists the models, gets a free quote, tells you the price, makes it on your prepaid balance once you say yes, and hands you the download link.

#### Is Revana part of Kie.ai?

No. Revana is a separate company, not affiliated with Kie.ai. This page compares the two APIs' public list prices.

Sources: each row's link; Revana prices https://revana.si/pricing; raw figures https://revana.si/data/market-prices.json; page https://revana.si/vs/kie

## Revana vs fal

Revana's list price is below fal's for 37 settings across 12 models: Seedance 2.5, Wan 3.0, Nano Banana Pro, Nano Banana 2, Veo 3.1 Fast, Veo 3.1, Seedance 2.0, Wan 2.6, Wan 2.7, Seedance 2.0 Fast, PixVerse V6 and Grok Imagine Video (checked 2026-10-30). fal lists the same price or less for some settings of Kling 3.0, Nano Banana 2, Veo 3.1 Fast and Veo 3.1.

- Seedance 2.5: $0.09 a second at 480p on Revana, against $0.2205 on fal (59% less).
- Wan 3.0: $0.07 a second at 720p on Revana, against $0.10 on fal (30% less).
- Nano Banana Pro: $0.04 an image at 1K on Revana, against $0.15 on fal (73% less).
- Nano Banana 2: $0.03 an image at 2K on Revana, against $0.12 on fal (75% less).
- Veo 3.1 Fast: $0.10 a second at 720p with sound on Revana, against $0.15 on fal (33% less).

### Popular models at a glance

- Seedance 2.5: Revana is lower at 480p ($0.09 against $0.2205, 59% less), at 720p ($0.21 against $0.473, 56% less) and at 1080p ($0.52 against $1.164, 55% less).
- Veo 3.1: Revana is lower at 4K with sound ($0.58 against $0.60, 3% less). Same price at 720p, 720p with sound, 1080p, 1080p with sound and 4K.
- Veo 3.1 Fast: Revana is lower at 720p ($0.08 against $0.10, 20% less), at 720p with sound ($0.10 against $0.15, 33% less) and at 1080p with sound ($0.11 against $0.15, 27% less), and at 2 more settings in the tables below. Same price at 1080p.
- Kling 3.0: fal is cheaper.
- Wan 3.0: Revana is lower at 480p ($0.04 against $0.05, 20% less), at 720p ($0.07 against $0.10, 30% less) and at 1080p ($0.14 against $0.20, 30% less).
- Nano Banana Pro: Revana is lower at 1K ($0.04 against $0.15, 73% less), at 2K ($0.04 against $0.15, 73% less) and at 4K ($0.08 against $0.30, 73% less).

### Where fal is cheaper

fal is cheaper than Revana on Kling 3.0 and Nano Banana 2, for at least one setting. It costs the same as Revana on some settings of Veo 3.1 and Veo 3.1 Fast.

### Video models per second (checked 2026-10-30, 2026-11-10)

| Model | Setting | Revana | fal | Revana is | Source |
|---|---|---|---|---|---|
| Seedance 2.5 | 480p | $0.09 | $0.2205 | 59% less | https://fal.ai/models/bytedance/seedance-2.5/text-to-video |
| Seedance 2.5 | 720p | $0.21 | $0.473 | 56% less | https://fal.ai/models/bytedance/seedance-2.5/text-to-video |
| Seedance 2.5 | 1080p | $0.52 | $1.164 | 55% less | https://fal.ai/models/bytedance/seedance-2.5/text-to-video |
| Wan 3.0 | 480p | $0.04 | $0.05 | 20% less | https://fal.ai/models/alibaba/wan-3.0/text-to-video |
| Wan 3.0 | 720p | $0.07 | $0.10 | 30% less | https://fal.ai/models/alibaba/wan-3.0/text-to-video |
| Wan 3.0 | 1080p | $0.14 | $0.20 | 30% less | https://fal.ai/models/alibaba/wan-3.0/text-to-video |
| Veo 3.1 Fast | 720p | $0.08 | $0.10 | 20% less | https://fal.ai/models/fal-ai/veo3.1/fast |
| Veo 3.1 Fast | 720p with sound | $0.10 | $0.15 | 33% less | https://fal.ai/models/fal-ai/veo3.1/fast |
| Veo 3.1 Fast | 1080p with sound | $0.11 | $0.15 | 27% less | https://fal.ai/models/fal-ai/veo3.1/fast |
| Veo 3.1 Fast | 4K | $0.25 | $0.30 | 17% less | https://fal.ai/models/fal-ai/veo3.1/fast |
| Veo 3.1 Fast | 4K with sound | $0.29 | $0.35 | 17% less | https://fal.ai/models/fal-ai/veo3.1/fast |
| Veo 3.1 | 4K with sound | $0.58 | $0.60 | 3% less | https://fal.ai/models/fal-ai/veo3.1 |
| Seedance 2.0 | 480p | $0.08 | $0.1345 | 41% less | https://fal.ai/models/bytedance/seedance-2.0/text-to-video |
| Seedance 2.0 | 720p | $0.15 | $0.3034 | 51% less | https://fal.ai/models/bytedance/seedance-2.0/text-to-video |
| Seedance 2.0 | 1080p | $0.40 | $0.682 | 41% less | https://fal.ai/models/bytedance/seedance-2.0/text-to-video |
| Wan 2.6 | 720p | $0.07 | $0.10 | 30% less | https://fal.ai/models/wan/v2.6/text-to-video |
| Wan 2.6 | 1080p | $0.11 | $0.15 | 27% less | https://fal.ai/models/wan/v2.6/text-to-video |
| Wan 2.7 | 720p | $0.09 | $0.10 | 10% less | https://fal.ai/models/fal-ai/wan/v2.7/text-to-video |
| Wan 2.7 | 1080p | $0.14 | $0.15 | 7% less | https://fal.ai/models/fal-ai/wan/v2.7/text-to-video |
| Seedance 2.0 Fast | 480p | $0.05 | $0.1076 | 54% less | https://fal.ai/models/bytedance/seedance-2.0/fast/text-to-video |
| Seedance 2.0 Fast | 720p | $0.11 | $0.2419 | 55% less | https://fal.ai/models/bytedance/seedance-2.0/fast/text-to-video |
| PixVerse V6 | 360p | $0.02 | $0.025 | 20% less | https://fal.ai/models/fal-ai/pixverse/v6/text-to-video |
| PixVerse V6 | 360p with sound | $0.03 | $0.035 | 14% less | https://fal.ai/models/fal-ai/pixverse/v6/text-to-video |
| PixVerse V6 | 540p | $0.03 | $0.035 | 14% less | https://fal.ai/models/fal-ai/pixverse/v6/text-to-video |
| PixVerse V6 | 540p with sound | $0.04 | $0.045 | 11% less | https://fal.ai/models/fal-ai/pixverse/v6/text-to-video |
| PixVerse V6 | 720p | $0.04 | $0.045 | 11% less | https://fal.ai/models/fal-ai/pixverse/v6/text-to-video |
| PixVerse V6 | 720p with sound | $0.05 | $0.06 | 17% less | https://fal.ai/models/fal-ai/pixverse/v6/text-to-video |
| PixVerse V6 | 1080p | $0.08 | $0.09 | 11% less | https://fal.ai/models/fal-ai/pixverse/v6/text-to-video |
| PixVerse V6 | 1080p with sound | $0.10 | $0.115 | 13% less | https://fal.ai/models/fal-ai/pixverse/v6/text-to-video |
| Grok Imagine Video | 480p | $0.02 | $0.05 | 60% less | https://fal.ai/models/xai/grok-imagine-video/text-to-video |
| Grok Imagine Video | 720p | $0.03 | $0.07 | 57% less | https://fal.ai/models/xai/grok-imagine-video/text-to-video |

### Image models per image (checked 2026-10-30)

| Model | Setting | Revana | fal | Revana is | Source |
|---|---|---|---|---|---|
| Nano Banana Pro | 1K | $0.04 | $0.15 | 73% less | https://fal.ai/models/fal-ai/nano-banana-pro |
| Nano Banana Pro | 2K | $0.04 | $0.15 | 73% less | https://fal.ai/models/fal-ai/nano-banana-pro |
| Nano Banana Pro | 4K | $0.08 | $0.30 | 73% less | https://fal.ai/models/fal-ai/nano-banana-pro |
| Nano Banana 2 | 1K | $0.03 | $0.08 | 63% less | https://fal.ai/models/fal-ai/nano-banana-2 |
| Nano Banana 2 | 2K | $0.03 | $0.12 | 75% less | https://fal.ai/models/fal-ai/nano-banana-2 |
| Nano Banana 2 | 4K | $0.04 | $0.16 | 75% less | https://fal.ai/models/fal-ai/nano-banana-2 |

These are Revana's list prices, charged on the first route. For Seedance 2.5, Wan 3.0, Nano Banana Pro, Nano Banana 2, Seedance 2.0 Fast and Grok Imagine Video, a backup route can cost more if the first one fails; each model page lists every route's price, and max_cost_usd sets the most a job may cost.

### How Revana and fal differ

fal runs generative media models on its own serverless GPU platform, prices each model per output on its model page, and also hosts models that teams deploy themselves.

Revana sells 139 video, image and music models and 99 language models through one API key and one prepaid balance. The language models are behind an OpenAI-compatible chat completions endpoint.

Revana routes each model across several providers where it can: it tries the lowest-cost route for the model and setting first and, if that route fails, moves the job to the next one. A job that fails on every route costs nothing, and a prompt the model refuses is not charged.

Credit is prepaid in packs from $10, with no subscription, seats or monthly minimum. Every finished job reports cost_usd, the exact amount charged.

An MCP server lets agents in Claude, ChatGPT, Cursor and Codex price and run generations directly.

### Use it from Claude, ChatGPT or Cursor

Revana has an MCP server, so Claude, ChatGPT, Cursor, VS Code and Codex can use it directly: add https://revana.si/api/mcp as a connector, approve it on revana.si, and ask for a video or an image. The agent lists the models, gets a free quote, tells you the price, makes it on your prepaid balance once you say yes, and hands you the download link.

### Questions

#### Is Revana cheaper than fal?

For Seedance 2.5, Wan 3.0, Nano Banana Pro, Nano Banana 2, Veo 3.1 Fast, Veo 3.1, Seedance 2.0, Wan 2.6, Wan 2.7, Seedance 2.0 Fast, PixVerse V6 and Grok Imagine Video, yes: Seedance 2.5: $0.09 a second at 480p on Revana, against $0.2205 on fal (59% less), checked 2026-11-10. Not for every model: fal lists the same price or less for some settings of Kling 3.0, Nano Banana 2, Veo 3.1 Fast and Veo 3.1.

#### What is a cheaper fal alternative for Seedance 2.5, Wan 3.0 and Nano Banana Pro?

Revana. Seedance 2.5: $0.09 a second at 480p on Revana, against $0.2205 on fal (59% less). Wan 3.0: $0.07 a second at 720p on Revana, against $0.10 on fal (30% less). Nano Banana Pro: $0.04 an image at 1K on Revana, against $0.15 on fal (73% less). Each price was checked on 2026-10-30 or later.

#### Am I charged for failed generations on Revana?

No. Revana holds the price when you submit and releases it if the job fails, including when the model's safety filter refuses the prompt. If a route fails, the router moves the job to the next route for that model; if none succeeds, the job costs nothing.

#### What happens on Revana when a provider is down?

Revana sends each job to the lowest-cost route for that model and setting and, if that route fails, moves it to the next one, so an outage at one provider does not stop the job where the model has another route. Of the models above, Seedance 2.5, Wan 3.0, Nano Banana Pro, Nano Banana 2, Veo 3.1 Fast, Veo 3.1, Seedance 2.0, Wan 2.6, Wan 2.7, Seedance 2.0 Fast and PixVerse V6 have more than one route.

#### Where is fal cheaper than Revana?

fal is cheaper than Revana on Kling 3.0 and Nano Banana 2, for at least one setting. It costs the same as Revana on some settings of Veo 3.1 and Veo 3.1 Fast. Every model is listed with its source in the raw price file, so you can check both price lists before moving.

#### Can I use Revana from Claude or ChatGPT?

Revana has an MCP server, so Claude, ChatGPT, Cursor, VS Code and Codex can use it directly: add https://revana.si/api/mcp as a connector, approve it on revana.si, and ask for a video or an image. The agent lists the models, gets a free quote, tells you the price, makes it on your prepaid balance once you say yes, and hands you the download link.

#### Is Revana part of fal?

No. Revana is a separate company, not affiliated with fal. This page compares the two APIs' public list prices.

Sources: each row's link; Revana prices https://revana.si/pricing; raw figures https://revana.si/data/market-prices.json; page https://revana.si/vs/fal

## Revana vs Replicate

Revana's list price is below Replicate's for 43 settings across 16 models: Seedance 2.5, Nano Banana Pro, Nano Banana 2, Veo 3.1 Fast, Kling 3.0, Seedance 2.0, Wan 2.6, Seedream 5.0 Pro, Ideogram V3, Seedance 2.0 Fast, Seedance 2.0 Mini, PixVerse V6, Grok Imagine Video, Grok Imagine Video 1.5, Flux 2 Flex and Seedream 5.0 Lite (checked 2026-10-30). Replicate lists the same price or less for some settings of 5 models, including Veo 3.1, Flux Kontext Pro, Veo 3.1 Fast and Seedance 2.0.

- Seedance 2.5: $0.09 a second at 480p on Revana, against $0.1028 on Replicate (12% less).
- Nano Banana Pro: $0.04 an image at 1K on Revana, against $0.15 on Replicate (73% less).
- Nano Banana 2: $0.04 an image at 4K on Revana, against $0.151 on Replicate (74% less).
- Veo 3.1 Fast: $0.10 a second at 720p with sound on Revana, against $0.15 on Replicate (33% less).
- Kling 3.0: $0.17 a second at 1080p with sound on Revana, against $0.336 on Replicate (49% less).

### Popular models at a glance

- Seedance 2.5: Revana is lower at 480p ($0.09 against $0.1028, 12% less) and at 720p ($0.21 against $0.2312, 9% less).
- Veo 3.1: Same price at 720p, 720p with sound, 1080p and 1080p with sound.
- Veo 3.1 Fast: Revana is lower at 720p ($0.08 against $0.10, 20% less), at 720p with sound ($0.10 against $0.15, 33% less) and at 1080p with sound ($0.11 against $0.15, 27% less). Same price at 1080p.
- Kling 3.0: Revana is lower at 1080p with sound ($0.17 against $0.336, 49% less).
- Wan 3.0: No Replicate price on file to compare.
- Nano Banana Pro: Revana is lower at 1K ($0.04 against $0.15, 73% less), at 2K ($0.04 against $0.15, 73% less) and at 4K ($0.08 against $0.30, 73% less).

### Where Replicate is cheaper

Replicate is not cheaper than Revana on any model compared here, and costs the same on some settings of Veo 3.1, Veo 3.1 Fast, Seedance 2.0, Ideogram V3 and Flux Kontext Pro.

### Video models per second (checked 2026-10-30, 2026-11-05, 2026-11-10)

| Model | Setting | Revana | Replicate | Revana is | Source |
|---|---|---|---|---|---|
| Seedance 2.5 | 480p | $0.09 | $0.1028 | 12% less | https://replicate.com/bytedance/seedance-2.5 |
| Seedance 2.5 | 720p | $0.21 | $0.2312 | 9% less | https://replicate.com/bytedance/seedance-2.5 |
| Veo 3.1 Fast | 720p | $0.08 | $0.10 | 20% less | https://replicate.com/google/veo-3.1-fast |
| Veo 3.1 Fast | 720p with sound | $0.10 | $0.15 | 33% less | https://replicate.com/google/veo-3.1-fast |
| Veo 3.1 Fast | 1080p with sound | $0.11 | $0.15 | 27% less | https://replicate.com/google/veo-3.1-fast |
| Kling 3.0 | 1080p with sound | $0.17 | $0.336 | 49% less | https://replicate.com/kwaivgi/kling-v3-video |
| Seedance 2.0 | 720p | $0.15 | $0.18 | 17% less | https://replicate.com/bytedance/seedance-2.0 |
| Seedance 2.0 | 1080p | $0.40 | $0.45 | 11% less | https://replicate.com/bytedance/seedance-2.0 |
| Seedance 2.0 | 4K | $0.83 | $1.00 | 17% less | https://replicate.com/bytedance/seedance-2.0 |
| Wan 2.6 | 720p | $0.07 | $0.10 | 30% less | https://replicate.com/wan-video/wan-2.6-t2v |
| Wan 2.6 | 1080p | $0.11 | $0.15 | 27% less | https://replicate.com/wan-video/wan-2.6-t2v |
| Seedance 2.0 Fast | 480p | $0.05 | $0.07 | 29% less | https://replicate.com/bytedance/seedance-2.0-fast |
| Seedance 2.0 Fast | 720p | $0.11 | $0.15 | 27% less | https://replicate.com/bytedance/seedance-2.0-fast |
| Seedance 2.0 Mini | 480p | $0.03 | $0.04 | 25% less | https://replicate.com/bytedance/seedance-2.0-mini |
| Seedance 2.0 Mini | 720p | $0.05 | $0.09 | 44% less | https://replicate.com/bytedance/seedance-2.0-mini |
| PixVerse V6 | 360p | $0.02 | $0.05 | 60% less | https://replicate.com/pixverse/pixverse-v6 |
| PixVerse V6 | 360p with sound | $0.03 | $0.07 | 57% less | https://replicate.com/pixverse/pixverse-v6 |
| PixVerse V6 | 540p | $0.03 | $0.07 | 57% less | https://replicate.com/pixverse/pixverse-v6 |
| PixVerse V6 | 540p with sound | $0.04 | $0.09 | 56% less | https://replicate.com/pixverse/pixverse-v6 |
| PixVerse V6 | 720p | $0.04 | $0.09 | 56% less | https://replicate.com/pixverse/pixverse-v6 |
| PixVerse V6 | 720p with sound | $0.05 | $0.12 | 58% less | https://replicate.com/pixverse/pixverse-v6 |
| PixVerse V6 | 1080p | $0.08 | $0.18 | 56% less | https://replicate.com/pixverse/pixverse-v6 |
| PixVerse V6 | 1080p with sound | $0.10 | $0.23 | 57% less | https://replicate.com/pixverse/pixverse-v6 |
| Grok Imagine Video | 480p | $0.02 | $0.05 | 60% less | https://replicate.com/xai/grok-imagine-video |
| Grok Imagine Video | 720p | $0.03 | $0.05 | 40% less | https://replicate.com/xai/grok-imagine-video |
| Grok Imagine Video 1.5 | 480p | $0.04 | $0.08 | 50% less | https://replicate.com/xai/grok-imagine-video-1.5 |
| Grok Imagine Video 1.5 | 720p | $0.07 | $0.08 | 12% less | https://replicate.com/xai/grok-imagine-video-1.5 |

### Image models per image (checked 2026-10-30)

| Model | Setting | Revana | Replicate | Revana is | Source |
|---|---|---|---|---|---|
| Nano Banana Pro | 1K | $0.04 | $0.15 | 73% less | https://replicate.com/google/nano-banana-pro |
| Nano Banana Pro | 2K | $0.04 | $0.15 | 73% less | https://replicate.com/google/nano-banana-pro |
| Nano Banana Pro | 4K | $0.08 | $0.30 | 73% less | https://replicate.com/google/nano-banana-pro |
| Nano Banana 2 | 1K | $0.03 | $0.067 | 55% less | https://replicate.com/google/nano-banana-2 |
| Nano Banana 2 | 2K | $0.03 | $0.101 | 70% less | https://replicate.com/google/nano-banana-2 |
| Nano Banana 2 | 4K | $0.04 | $0.151 | 74% less | https://replicate.com/google/nano-banana-2 |
| Seedream 5.0 Pro | 1K | $0.04 | $0.045 | 11% less | https://replicate.com/bytedance/seedream-5-pro |
| Seedream 5.0 Pro | 2K | $0.04 | $0.09 | 56% less | https://replicate.com/bytedance/seedream-5-pro |
| Ideogram V3 | balanced | $0.05 | $0.06 | 17% less | https://replicate.com/ideogram-ai/ideogram-v3-balanced |
| Ideogram V3 | quality | $0.07 | $0.09 | 22% less | https://replicate.com/ideogram-ai/ideogram-v3-quality |
| Flux 2 Flex | 1MP | $0.05 | $0.06 | 17% less | https://replicate.com/black-forest-labs/flux-2-flex |
| Flux 2 Flex | 2MP | $0.10 | $0.12 | 17% less | https://replicate.com/black-forest-labs/flux-2-flex |
| Flux 2 Flex | 3MP | $0.15 | $0.18 | 17% less | https://replicate.com/black-forest-labs/flux-2-flex |
| Flux 2 Flex | 4MP | $0.20 | $0.24 | 17% less | https://replicate.com/black-forest-labs/flux-2-flex |
| Seedream 5.0 Lite | 2K | $0.03 | $0.035 | 14% less | https://replicate.com/bytedance/seedream-5-lite |
| Seedream 5.0 Lite | 3K | $0.03 | $0.035 | 14% less | https://replicate.com/bytedance/seedream-5-lite |

These are Revana's list prices, charged on the first route. For Seedance 2.5, Nano Banana Pro, Nano Banana 2, Seedream 5.0 Pro, Seedance 2.0 Fast, Seedance 2.0 Mini, Grok Imagine Video, Grok Imagine Video 1.5 and Seedream 5.0 Lite, a backup route can cost more if the first one fails; each model page lists every route's price, and max_cost_usd sets the most a job may cost.

### How Revana and Replicate differ

Replicate runs official and open-source models behind one API, priced per output or per second of hardware, and lets you publish your own models.

Revana sells 139 video, image and music models and 99 language models through one API key and one prepaid balance. The language models are behind an OpenAI-compatible chat completions endpoint.

Revana routes each model across several providers where it can: it tries the lowest-cost route for the model and setting first and, if that route fails, moves the job to the next one. A job that fails on every route costs nothing, and a prompt the model refuses is not charged.

Credit is prepaid in packs from $10, with no subscription, seats or monthly minimum. Every finished job reports cost_usd, the exact amount charged.

An MCP server lets agents in Claude, ChatGPT, Cursor and Codex price and run generations directly.

### Use it from Claude, ChatGPT or Cursor

Revana has an MCP server, so Claude, ChatGPT, Cursor, VS Code and Codex can use it directly: add https://revana.si/api/mcp as a connector, approve it on revana.si, and ask for a video or an image. The agent lists the models, gets a free quote, tells you the price, makes it on your prepaid balance once you say yes, and hands you the download link.

### Questions

#### Is Revana cheaper than Replicate?

For Seedance 2.5, Nano Banana Pro, Nano Banana 2, Veo 3.1 Fast, Kling 3.0, Seedance 2.0, Wan 2.6, Seedream 5.0 Pro, Ideogram V3, Seedance 2.0 Fast, Seedance 2.0 Mini, PixVerse V6, Grok Imagine Video, Grok Imagine Video 1.5, Flux 2 Flex and Seedream 5.0 Lite, yes: Seedance 2.5: $0.09 a second at 480p on Revana, against $0.1028 on Replicate (12% less), checked 2026-11-10. Not for every model: Replicate lists the same price or less for some settings of 5 models, including Veo 3.1, Flux Kontext Pro, Veo 3.1 Fast and Seedance 2.0.

#### What is a cheaper Replicate alternative for Seedance 2.5, Nano Banana Pro and Nano Banana 2?

Revana. Seedance 2.5: $0.09 a second at 480p on Revana, against $0.1028 on Replicate (12% less). Nano Banana Pro: $0.04 an image at 1K on Revana, against $0.15 on Replicate (73% less). Nano Banana 2: $0.04 an image at 4K on Revana, against $0.151 on Replicate (74% less). Each price was checked on 2026-10-30 or later.

#### Am I charged for failed generations on Revana?

No. Revana holds the price when you submit and releases it if the job fails, including when the model's safety filter refuses the prompt. If a route fails, the router moves the job to the next route for that model; if none succeeds, the job costs nothing.

#### What happens on Revana when a provider is down?

Revana sends each job to the lowest-cost route for that model and setting and, if that route fails, moves it to the next one, so an outage at one provider does not stop the job where the model has another route. Of the models above, Seedance 2.5, Nano Banana Pro, Nano Banana 2, Veo 3.1 Fast, Kling 3.0, Seedance 2.0, Wan 2.6, Seedream 5.0 Pro, Seedance 2.0 Fast, Seedance 2.0 Mini, PixVerse V6 and Seedream 5.0 Lite have more than one route.

#### Where is Replicate cheaper than Revana?

Replicate is not cheaper than Revana on any model compared here, and costs the same on some settings of Veo 3.1, Veo 3.1 Fast, Seedance 2.0, Ideogram V3 and Flux Kontext Pro. Every model is listed with its source in the raw price file, so you can check both price lists before moving.

#### Can I use Revana from Claude or ChatGPT?

Revana has an MCP server, so Claude, ChatGPT, Cursor, VS Code and Codex can use it directly: add https://revana.si/api/mcp as a connector, approve it on revana.si, and ask for a video or an image. The agent lists the models, gets a free quote, tells you the price, makes it on your prepaid balance once you say yes, and hands you the download link.

#### Is Revana part of Replicate?

No. Revana is a separate company, not affiliated with Replicate. This page compares the two APIs' public list prices.

Sources: each row's link; Revana prices https://revana.si/pricing; raw figures https://revana.si/data/market-prices.json; page https://revana.si/vs/replicate

# Every language model

## Claude Fable 5 API

> The previous Fable, for long-running agentic work. By Anthropic, on Revana at $3.50 in / $17.50 out per 1M tokens.

- Model id: `claude-fable-5`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/claude-fable-5

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $3.50 |
| Cached input | $0.35 |
| Cache write (5 min) | $4.375 |
| Cache write (1 hour) | $7.00 |
| Output | $17.50 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-fable-5","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Fable 5.1 API

> Anthropic's model for demanding reasoning and long-horizon agentic work. By Anthropic, on Revana at $3.50 in / $17.50 out per 1M tokens.

- Model id: `claude-fable-5-1`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/claude-fable-5-1

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $3.50 |
| Cached input | $0.0875 |
| Cache write (5 min) | $4.375 |
| Cache write (1 hour) | $7.00 |
| Output | $17.50 |

A request at reasoning effort max is billed at 3× these rates.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-fable-5-1","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Haiku 4.5 API

> Claude Haiku 4.5 by Anthropic, through an OpenAI-compatible chat completions API. By Anthropic, on Revana at $1.00 in / $5.00 out per 1M tokens.

- Model id: `claude-haiku-4-5-20251001`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/claude-haiku-4-5-20251001

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.00 |
| Cached input | $0.10 |
| Cache write (5 min) | $1.25 |
| Cache write (1 hour) | $2.00 |
| Output | $5.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-haiku-4-5-20251001","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Haiku 4.5 (thinking) API

> Claude Haiku 4.5 (thinking) by Anthropic, through an OpenAI-compatible chat completions API. By Anthropic, on Revana at $1.00 in / $5.00 out per 1M tokens.

- Model id: `claude-haiku-4-5-20251001-thinking`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/claude-haiku-4-5-20251001-thinking

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.00 |
| Cached input | $1.00 |
| Cache write (5 min) | $1.00 |
| Cache write (1 hour) | $1.00 |
| Output | $5.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-haiku-4-5-20251001-thinking","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Haiku 5.5 API

> For high-volume, latency-sensitive tasks such as classification, extraction and routing. By Anthropic, on Revana at $0.035 in / $0.175 out per 1M tokens.

- Model id: `claude-haiku-5-5`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/claude-haiku-5-5

### Price (USD per million tokens)

| Tokens | Price | Prompt over 100K |
|---|---|---|
| Input | $0.035 | $0.175 |
| Cached input | $0.0035 | $0.0175 |
| Cache write (5 min) | $0.0438 | $0.2188 |
| Cache write (1 hour) | $0.07 | $0.35 |
| Output | $0.175 | $0.875 |

A request whose prompt is over 100K tokens is billed at $0.175 input and $0.875 output per million tokens.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-haiku-5-5","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Opus 4.5 API

> Claude Opus 4.5 by Anthropic, through an OpenAI-compatible chat completions API. By Anthropic, on Revana at $5.00 in / $25.00 out per 1M tokens.

- Model id: `claude-opus-4-5-20251101`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/claude-opus-4-5-20251101

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $5.00 |
| Cached input | $5.00 |
| Cache write (5 min) | $5.00 |
| Cache write (1 hour) | $5.00 |
| Output | $25.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-opus-4-5-20251101","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Opus 4.5 (thinking) API

> Claude Opus 4.5 (thinking) by Anthropic, through an OpenAI-compatible chat completions API. By Anthropic, on Revana at $5.00 in / $25.00 out per 1M tokens.

- Model id: `claude-opus-4-5-20251101-thinking`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/claude-opus-4-5-20251101-thinking

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $5.00 |
| Cached input | $5.00 |
| Cache write (5 min) | $5.00 |
| Cache write (1 hour) | $5.00 |
| Output | $25.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-opus-4-5-20251101-thinking","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Opus 4.6 API

> Claude Opus 4.6 by Anthropic, through an OpenAI-compatible chat completions API. By Anthropic, on Revana at $5.00 in / $10.00 out per 1M tokens.

- Model id: `claude-opus-4-6`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 1M tokens
- Routes: 1
- Page: https://revana.si/llm/claude-opus-4-6

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $5.00 |
| Cached input | $0.50 |
| Cache write (5 min) | $6.25 |
| Cache write (1 hour) | $6.25 |
| Output | $10.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-opus-4-6","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Opus 4.6 (thinking) API

> Claude Opus 4.6 (thinking) by Anthropic, through an OpenAI-compatible chat completions API. By Anthropic, on Revana at $5.00 in / $25.00 out per 1M tokens.

- Model id: `claude-opus-4-6-thinking`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/claude-opus-4-6-thinking

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $5.00 |
| Cached input | $5.00 |
| Cache write (5 min) | $5.00 |
| Cache write (1 hour) | $5.00 |
| Output | $25.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-opus-4-6-thinking","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Opus 4.7 API

> Claude Opus 4.7 by Anthropic, through an OpenAI-compatible chat completions API. By Anthropic, on Revana at $5.00 in / $25.00 out per 1M tokens.

- Model id: `claude-opus-4-7`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/claude-opus-4-7

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $5.00 |
| Cached input | $0.50 |
| Cache write (5 min) | $6.25 |
| Cache write (1 hour) | $10.00 |
| Output | $25.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-opus-4-7","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Opus 4.8 API

> The previous Opus generation, for complex coding and agent work. By Anthropic, on Revana at $1.75 in / $8.75 out per 1M tokens.

- Model id: `claude-opus-4-8`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/claude-opus-4-8

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.75 |
| Cached input | $0.175 |
| Cache write (5 min) | $2.1875 |
| Cache write (1 hour) | $3.50 |
| Output | $8.75 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-opus-4-8","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Opus 5 API

> For complex agentic coding and enterprise work. By Anthropic, on Revana at $1.75 in / $8.75 out per 1M tokens.

- Model id: `claude-opus-5`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/claude-opus-5

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.75 |
| Cached input | $0.175 |
| Cache write (5 min) | $2.1875 |
| Cache write (1 hour) | $3.50 |
| Output | $8.75 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-opus-5","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Opus 5.5 API

> For long-running agentic coding and knowledge work. By Anthropic, on Revana at $1.40 in / $7.00 out per 1M tokens.

- Model id: `claude-opus-5-5`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/claude-opus-5-5

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.40 |
| Cached input | $0.07 |
| Cache write (5 min) | $1.75 |
| Cache write (1 hour) | $2.80 |
| Output | $7.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-opus-5-5","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Sonnet 4.5 API

> Claude Sonnet 4.5 by Anthropic, through an OpenAI-compatible chat completions API. By Anthropic, on Revana at $3.00 in / $15.00 out per 1M tokens.

- Model id: `claude-sonnet-4-5-20250929`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/claude-sonnet-4-5-20250929

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $3.00 |
| Cached input | $0.30 |
| Cache write (5 min) | $3.75 |
| Cache write (1 hour) | $6.00 |
| Output | $15.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-sonnet-4-5-20250929","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Sonnet 4.5 (thinking) API

> Claude Sonnet 4.5 (thinking) by Anthropic, through an OpenAI-compatible chat completions API. By Anthropic, on Revana at $3.00 in / $15.00 out per 1M tokens.

- Model id: `claude-sonnet-4-5-20250929-thinking`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/claude-sonnet-4-5-20250929-thinking

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $3.00 |
| Cached input | $3.00 |
| Cache write (5 min) | $3.00 |
| Cache write (1 hour) | $3.00 |
| Output | $15.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-sonnet-4-5-20250929-thinking","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Sonnet 4.6 API

> Claude Sonnet 4.6 by Anthropic, through an OpenAI-compatible chat completions API. By Anthropic, on Revana at $3.00 in / $15.00 out per 1M tokens.

- Model id: `claude-sonnet-4-6`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/claude-sonnet-4-6

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $3.00 |
| Cached input | $0.30 |
| Cache write (5 min) | $3.75 |
| Cache write (1 hour) | $6.00 |
| Output | $15.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-sonnet-4-6","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Sonnet 4.6 (thinking) API

> Claude Sonnet 4.6 (thinking) by Anthropic, through an OpenAI-compatible chat completions API. By Anthropic, on Revana at $3.00 in / $15.00 out per 1M tokens.

- Model id: `claude-sonnet-4-6-thinking`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/claude-sonnet-4-6-thinking

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $3.00 |
| Cached input | $3.00 |
| Cache write (5 min) | $3.00 |
| Cache write (1 hour) | $3.00 |
| Output | $15.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-sonnet-4-6-thinking","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Sonnet 5 API

> The best combination of speed and intelligence in the Claude lineup. By Anthropic, on Revana at $0.70 in / $3.50 out per 1M tokens.

- Model id: `claude-sonnet-5`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/claude-sonnet-5

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.70 |
| Cached input | $0.07 |
| Cache write (5 min) | $0.875 |
| Cache write (1 hour) | $1.40 |
| Output | $3.50 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-sonnet-5","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Claude Sonnet 5.5 API

> The best combination of speed and intelligence, as Anthropic's latest Sonnet. By Anthropic, on Revana at $0.70 in / $3.50 out per 1M tokens.

- Model id: `claude-sonnet-5-5`
- Kind: text (chat completions)
- Maker: Anthropic
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/claude-sonnet-5-5

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.70 |
| Cached input | $0.035 |
| Cache write (5 min) | $0.875 |
| Cache write (1 hour) | $1.40 |
| Output | $3.50 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-sonnet-5-5","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## DeepSeek R1 API

> DeepSeek R1 by DeepSeek, through an OpenAI-compatible chat completions API. By DeepSeek, on Revana at $0.56 in / $2.24 out per 1M tokens.

- Model id: `deepseek-r1`
- Kind: text (chat completions)
- Maker: DeepSeek
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/deepseek-r1

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.56 |
| Cached input | $0.56 |
| Cache write (5 min) | $0.56 |
| Cache write (1 hour) | $0.56 |
| Output | $2.24 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"deepseek-r1","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## DeepSeek V3.2 API

> DeepSeek V3.2 by DeepSeek, through an OpenAI-compatible chat completions API. By DeepSeek, on Revana at $0.257 in / $0.3855 out per 1M tokens.

- Model id: `deepseek-v3.2`
- Kind: text (chat completions)
- Maker: DeepSeek
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/deepseek-v3.2

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.257 |
| Cached input | $0.257 |
| Cache write (5 min) | $0.257 |
| Cache write (1 hour) | $0.257 |
| Output | $0.3855 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"deepseek-v3.2","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## DeepSeek V4 Flash API

> DeepSeek V4 Flash by DeepSeek, through an OpenAI-compatible chat completions API. By DeepSeek, on Revana at $0.4286 in / $1.2857 out per 1M tokens.

- Model id: `deepseek-v4-flash`
- Kind: text (chat completions)
- Maker: DeepSeek
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 393K tokens
- Routes: 1
- Page: https://revana.si/llm/deepseek-v4-flash

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.4286 |
| Cached input | $0.0429 |
| Cache write (5 min) | $0.4286 |
| Cache write (1 hour) | $0.4286 |
| Output | $1.2857 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"deepseek-v4-flash","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## DeepSeek V4 Pro API

> DeepSeek's strongest model, upgraded for agent work. By DeepSeek, on Revana at $0.396 in / $1.188 out per 1M tokens.

- Model id: `deepseek-v4-pro`
- Kind: text (chat completions)
- Maker: DeepSeek
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 393K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/deepseek-v4-pro

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.396 |
| Cached input | $0.0132 |
| Cache write (5 min) | $0.396 |
| Cache write (1 hour) | $0.792 |
| Output | $1.188 |

Requests made weekdays 09:00-12:00 and 14:00-18:00 Beijing time (UTC+8) are billed at 2× these rates.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"deepseek-v4-pro","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## DeepSeek V4.1 Flash API

> DeepSeek's fast, low-cost model, with native image understanding. By DeepSeek, on Revana at $0.09 in / $0.36 out per 1M tokens.

- Model id: `deepseek-v4.1-flash`
- Kind: text (chat completions)
- Maker: DeepSeek
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 393K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/deepseek-v4.1-flash

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.09 |
| Cached input | $0.0018 |
| Cache write (5 min) | $0.09 |
| Cache write (1 hour) | $0.18 |
| Output | $0.36 |

Requests made weekdays 09:00-12:00 and 14:00-18:00 Beijing time (UTC+8) are billed at 2× these rates.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"deepseek-v4.1-flash","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Gemini 3 Flash Preview API

> Gemini 3 Flash Preview by Google, through an OpenAI-compatible chat completions API. By Google, on Revana at $0.50 in / $3.00 out per 1M tokens.

- Model id: `gemini-3-flash-preview`
- Kind: text (chat completions)
- Maker: Google
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/gemini-3-flash-preview

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.50 |
| Cached input | $0.50 |
| Cache write (5 min) | $0.50 |
| Cache write (1 hour) | $0.50 |
| Output | $3.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-3-flash-preview","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Gemini 3 Flash Preview (no thinking) API

> Gemini 3 Flash Preview (no thinking) by Google, through an OpenAI-compatible chat completions API. By Google, on Revana at $0.50 in / $3.00 out per 1M tokens.

- Model id: `gemini-3-flash-preview-nothinking`
- Kind: text (chat completions)
- Maker: Google
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/gemini-3-flash-preview-nothinking

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.50 |
| Cached input | $0.50 |
| Cache write (5 min) | $0.50 |
| Cache write (1 hour) | $0.50 |
| Output | $3.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-3-flash-preview-nothinking","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Gemini 3 Flash Preview (thinking) API

> Gemini 3 Flash Preview (thinking) by Google, through an OpenAI-compatible chat completions API. By Google, on Revana at $0.50 in / $3.00 out per 1M tokens.

- Model id: `gemini-3-flash-preview-thinking`
- Kind: text (chat completions)
- Maker: Google
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/gemini-3-flash-preview-thinking

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.50 |
| Cached input | $0.50 |
| Cache write (5 min) | $0.50 |
| Cache write (1 hour) | $0.50 |
| Output | $3.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-3-flash-preview-thinking","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Gemini 3 Pro Preview API

> Gemini 3 Pro Preview by Google, through an OpenAI-compatible chat completions API. By Google, on Revana at $2.00 in / $12.00 out per 1M tokens.

- Model id: `gemini-3-pro-preview`
- Kind: text (chat completions)
- Maker: Google
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/gemini-3-pro-preview

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $2.00 |
| Cached input | $2.00 |
| Cache write (5 min) | $2.00 |
| Cache write (1 hour) | $2.00 |
| Output | $12.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-3-pro-preview","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Gemini 3 Pro Preview (thinking) API

> Gemini 3 Pro Preview (thinking) by Google, through an OpenAI-compatible chat completions API. By Google, on Revana at $2.00 in / $12.00 out per 1M tokens.

- Model id: `gemini-3-pro-preview-thinking`
- Kind: text (chat completions)
- Maker: Google
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/gemini-3-pro-preview-thinking

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $2.00 |
| Cached input | $2.00 |
| Cache write (5 min) | $2.00 |
| Cache write (1 hour) | $2.00 |
| Output | $12.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-3-pro-preview-thinking","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Gemini 3.1 Flash Lite Preview API

> Gemini 3.1 Flash Lite Preview by Google, through an OpenAI-compatible chat completions API. By Google, on Revana at $0.25 in / $1.50 out per 1M tokens.

- Model id: `gemini-3.1-flash-lite-preview`
- Kind: text (chat completions)
- Maker: Google
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/gemini-3.1-flash-lite-preview

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.25 |
| Cached input | $0.25 |
| Cache write (5 min) | $0.25 |
| Cache write (1 hour) | $0.25 |
| Output | $1.50 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-3.1-flash-lite-preview","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Gemini 3.1 Pro Preview API

> Gemini 3.1 Pro Preview by Google, through an OpenAI-compatible chat completions API. By Google, on Revana at $2.00 in / $12.00 out per 1M tokens.

- Model id: `gemini-3.1-pro-preview`
- Kind: text (chat completions)
- Maker: Google
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/gemini-3.1-pro-preview

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $2.00 |
| Cached input | $2.00 |
| Cache write (5 min) | $2.00 |
| Cache write (1 hour) | $2.00 |
| Output | $12.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-3.1-pro-preview","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Gemini 3.1 Pro Preview (thinking) API

> Gemini 3.1 Pro Preview (thinking) by Google, through an OpenAI-compatible chat completions API. By Google, on Revana at $2.00 in / $12.00 out per 1M tokens.

- Model id: `gemini-3.1-pro-preview-thinking`
- Kind: text (chat completions)
- Maker: Google
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/gemini-3.1-pro-preview-thinking

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $2.00 |
| Cached input | $2.00 |
| Cache write (5 min) | $2.00 |
| Cache write (1 hour) | $2.00 |
| Output | $12.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-3.1-pro-preview-thinking","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Gemini 3.5 Flash API

> Gemini 3.5 Flash by Google, through an OpenAI-compatible chat completions API. By Google, on Revana at $1.50 in / $9.00 out per 1M tokens.

- Model id: `gemini-3.5-flash`
- Kind: text (chat completions)
- Maker: Google
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/gemini-3.5-flash

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.50 |
| Cached input | $1.50 |
| Cache write (5 min) | $1.50 |
| Cache write (1 hour) | $1.50 |
| Output | $9.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-3.5-flash","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Gemini 3.5 Flash Lite API

> Gemini 3.5 Flash Lite by Google, through an OpenAI-compatible chat completions API. By Google, on Revana at $0.30 in / $2.499 out per 1M tokens.

- Model id: `gemini-3.5-flash-lite`
- Kind: text (chat completions)
- Maker: Google
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/gemini-3.5-flash-lite

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.30 |
| Cached input | $0.30 |
| Cache write (5 min) | $0.30 |
| Cache write (1 hour) | $0.30 |
| Output | $2.499 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-3.5-flash-lite","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Gemini 3.6 Flash API

> Gemini 3.6 Flash by Google, through an OpenAI-compatible chat completions API. By Google, on Revana at $1.50 in / $7.50 out per 1M tokens.

- Model id: `gemini-3.6-flash`
- Kind: text (chat completions)
- Maker: Google
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1.05M tokens
- Max output: 66K tokens
- Routes: 1
- Page: https://revana.si/llm/gemini-3.6-flash

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.50 |
| Cached input | $0.15 |
| Cache write (5 min) | $1.50 |
| Cache write (1 hour) | $1.50 |
| Output | $7.50 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-3.6-flash","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Gemini 3.7 Flash API

> Gemini 3.7 Flash by Google, through an OpenAI-compatible chat completions API. By Google, on Revana at $0.75 in / $3.75 out per 1M tokens.

- Model id: `gemini-3.7-flash`
- Kind: text (chat completions)
- Maker: Google
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1.05M tokens
- Max output: 66K tokens
- Routes: 1
- Page: https://revana.si/llm/gemini-3.7-flash

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.75 |
| Cached input | $0.075 |
| Cache write (5 min) | $0.75 |
| Cache write (1 hour) | $0.75 |
| Output | $3.75 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-3.7-flash","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Gemini 3.8 Flash API

> Google's fast, low-cost Gemini model for everyday text and image tasks. By Google, on Revana at $0.225 in / $1.125 out per 1M tokens.

- Model id: `gemini-3.8-flash`
- Kind: text (chat completions)
- Maker: Google
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 64K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/gemini-3.8-flash

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.225 |
| Cached input | $0.0225 |
| Cache write (5 min) | $0.225 |
| Cache write (1 hour) | $0.45 |
| Output | $1.125 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-3.8-flash","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Gemini 4 Argon API

> Google's flagship Gemini model for complex reasoning, coding and long-context work. By Google, on Revana at $0.60 in / $3.00 out per 1M tokens.

- Model id: `gemini-4-argon`
- Kind: text (chat completions)
- Maker: Google
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 64K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/gemini-4-argon

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.60 |
| Cached input | $0.03 |
| Cache write (5 min) | $0.60 |
| Cache write (1 hour) | $1.20 |
| Output | $3.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-4-argon","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Kimi K2.5 API

> Kimi K2.5 by Kimi, through an OpenAI-compatible chat completions API. By Kimi, on Revana at $0.514 in / $0.6939 out per 1M tokens.

- Model id: `kimi-k2.5`
- Kind: text (chat completions)
- Maker: Kimi
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/kimi-k2.5

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.514 |
| Cached input | $0.514 |
| Cache write (5 min) | $0.514 |
| Cache write (1 hour) | $0.514 |
| Output | $0.6939 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"kimi-k2.5","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Kimi K2.6 API

> Kimi's general model for chat and agent tasks. By Kimi, on Revana at $0.57 in / $2.40 out per 1M tokens.

- Model id: `kimi-k2.6`
- Kind: text (chat completions)
- Maker: Kimi
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 256K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/kimi-k2.6

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.57 |
| Cached input | $0.096 |
| Cache write (5 min) | $0.57 |
| Cache write (1 hour) | $1.14 |
| Output | $2.40 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"kimi-k2.6","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Kimi K2.7 Code API

> Kimi's dedicated coding model. By Kimi, on Revana at $0.57 in / $2.40 out per 1M tokens.

- Model id: `kimi-k2.7-code`
- Kind: text (chat completions)
- Maker: Kimi
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 256K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/kimi-k2.7-code

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.57 |
| Cached input | $0.114 |
| Cache write (5 min) | $0.57 |
| Cache write (1 hour) | $1.14 |
| Output | $2.40 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"kimi-k2.7-code","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Kimi K3 API

> Kimi's most capable model, with visual understanding. By Kimi, on Revana at $1.80 in / $9.00 out per 1M tokens.

- Model id: `kimi-k3`
- Kind: text (chat completions)
- Maker: Kimi
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 1M tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/kimi-k3

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.80 |
| Cached input | $0.18 |
| Cache write (5 min) | $1.80 |
| Cache write (1 hour) | $3.60 |
| Output | $9.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"kimi-k3","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Kimi K3.1 Pro API

> Kimi's flagship reasoning model, with visual understanding. By Kimi, on Revana at $1.20 in / $6.00 out per 1M tokens.

- Model id: `kimi-k3.1-pro`
- Kind: text (chat completions)
- Maker: Kimi
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 1M tokens
- Routes: 1
- Page: https://revana.si/llm/kimi-k3.1-pro

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.20 |
| Cached input | $0.12 |
| Cache write (5 min) | $1.20 |
| Cache write (1 hour) | $2.40 |
| Output | $6.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"kimi-k3.1-pro","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Kimi K3.1 Pro Highspeed API

> Kimi K3.1 Pro with 2-3x faster output. By Kimi, on Revana at $2.40 in / $12.00 out per 1M tokens.

- Model id: `kimi-k3.1-pro-highspeed`
- Kind: text (chat completions)
- Maker: Kimi
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 1M tokens
- Routes: 1
- Page: https://revana.si/llm/kimi-k3.1-pro-highspeed

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $2.40 |
| Cached input | $0.12 |
| Cache write (5 min) | $2.40 |
| Cache write (1 hour) | $4.80 |
| Output | $12.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"kimi-k3.1-pro-highspeed","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## MiniMax M2.5 API

> MiniMax M2.5 by MiniMax, through an OpenAI-compatible chat completions API. By MiniMax, on Revana at $0.30 in / $1.20 out per 1M tokens.

- Model id: `minimax-m2.5`
- Kind: text (chat completions)
- Maker: MiniMax
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/minimax-m2.5

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.30 |
| Cached input | $0.30 |
| Cache write (5 min) | $0.30 |
| Cache write (1 hour) | $0.30 |
| Output | $1.20 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"minimax-m2.5","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## MiniMax M3 API

> MiniMax M3 by MiniMax, through an OpenAI-compatible chat completions API. By MiniMax, on Revana at $0.30 in / $1.20 out per 1M tokens.

- Model id: `minimax-m3`
- Kind: text (chat completions)
- Maker: MiniMax
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/minimax-m3

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.30 |
| Cached input | $0.30 |
| Cache write (5 min) | $0.30 |
| Cache write (1 hour) | $0.30 |
| Output | $1.20 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"minimax-m3","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-4.1 API

> GPT-4.1 by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $2.00 in / $8.00 out per 1M tokens.

- Model id: `gpt-4.1`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-4.1

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $2.00 |
| Cached input | $0.50 |
| Cache write (5 min) | $2.00 |
| Cache write (1 hour) | $2.00 |
| Output | $8.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-4.1","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-4.1 mini API

> GPT-4.1 mini by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $0.40 in / $1.60 out per 1M tokens.

- Model id: `gpt-4.1-mini`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-4.1-mini

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.40 |
| Cached input | $0.10 |
| Cache write (5 min) | $0.40 |
| Cache write (1 hour) | $0.40 |
| Output | $1.60 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-4.1-mini","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-4.1 nano API

> GPT-4.1 nano by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $0.10 in / $0.40 out per 1M tokens.

- Model id: `gpt-4.1-nano`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-4.1-nano

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.10 |
| Cached input | $0.025 |
| Cache write (5 min) | $0.10 |
| Cache write (1 hour) | $0.10 |
| Output | $0.40 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-4.1-nano","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-4o API

> GPT-4o by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $2.50 in / $10.00 out per 1M tokens.

- Model id: `gpt-4o`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-4o

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $2.50 |
| Cached input | $1.25 |
| Cache write (5 min) | $2.50 |
| Cache write (1 hour) | $2.50 |
| Output | $10.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-4o","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-4o mini API

> GPT-4o mini by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $0.15 in / $0.60 out per 1M tokens.

- Model id: `gpt-4o-mini`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/gpt-4o-mini

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.15 |
| Cached input | $0.15 |
| Cache write (5 min) | $0.15 |
| Cache write (1 hour) | $0.15 |
| Output | $0.60 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-4o-mini","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-5 API

> GPT-5 by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $1.25 in / $10.00 out per 1M tokens.

- Model id: `gpt-5`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-5

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.25 |
| Cached input | $0.125 |
| Cache write (5 min) | $1.25 |
| Cache write (1 hour) | $1.25 |
| Output | $10.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-5 mini API

> GPT-5 mini by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $0.25 in / $2.00 out per 1M tokens.

- Model id: `gpt-5-mini`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-5-mini

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.25 |
| Cached input | $0.025 |
| Cache write (5 min) | $0.25 |
| Cache write (1 hour) | $0.25 |
| Output | $2.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5-mini","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-5 nano API

> GPT-5 nano by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $0.05 in / $0.40 out per 1M tokens.

- Model id: `gpt-5-nano`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-5-nano

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.05 |
| Cached input | $0.005 |
| Cache write (5 min) | $0.05 |
| Cache write (1 hour) | $0.05 |
| Output | $0.40 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5-nano","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-5.1 API

> GPT-5.1 by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $1.25 in / $10.00 out per 1M tokens.

- Model id: `gpt-5.1`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-5.1

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.25 |
| Cached input | $0.125 |
| Cache write (5 min) | $1.25 |
| Cache write (1 hour) | $1.25 |
| Output | $10.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.1","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-5.2 API

> GPT-5.2 by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $1.75 in / $14.00 out per 1M tokens.

- Model id: `gpt-5.2`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-5.2

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.75 |
| Cached input | $0.175 |
| Cache write (5 min) | $1.75 |
| Cache write (1 hour) | $1.75 |
| Output | $14.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.2","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-5.3 Codex API

> GPT-5.3 Codex by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $1.75 in / $14.00 out per 1M tokens.

- Model id: `gpt-5.3-codex`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/gpt-5.3-codex

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.75 |
| Cached input | $1.75 |
| Cache write (5 min) | $1.75 |
| Cache write (1 hour) | $1.75 |
| Output | $14.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.3-codex","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-5.4 API

> GPT-5.4 by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $2.50 in / $15.00 out per 1M tokens.

- Model id: `gpt-5.4`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 13K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-5.4

### Price (USD per million tokens)

| Tokens | Price | Prompt over 272K |
|---|---|---|
| Input | $2.50 | $5.00 |
| Cached input | $0.25 | $0.50 |
| Cache write (5 min) | $2.50 | $5.00 |
| Cache write (1 hour) | $2.50 | $5.00 |
| Output | $15.00 | $22.50 |

A request whose prompt is over 272K tokens is billed at $5.00 input and $22.50 output per million tokens.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.4","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-5.4 mini API

> GPT-5.4 mini by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $0.75 in / $4.50 out per 1M tokens.

- Model id: `gpt-5.4-mini`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-5.4-mini

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.75 |
| Cached input | $0.075 |
| Cache write (5 min) | $0.75 |
| Cache write (1 hour) | $0.75 |
| Output | $4.50 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.4-mini","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-5.4 nano API

> GPT-5.4 nano by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $0.20 in / $1.25 out per 1M tokens.

- Model id: `gpt-5.4-nano`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-5.4-nano

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.20 |
| Cached input | $0.02 |
| Cache write (5 min) | $0.20 |
| Cache write (1 hour) | $0.20 |
| Output | $1.25 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.4-nano","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-5.5 API

> A flagship for coding and the most complex professional work. By OpenAI, on Revana at $1.15 in / $6.90 out per 1M tokens.

- Model id: `gpt-5.5`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1.05M tokens
- Max output: 128K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-5.5

### Price (USD per million tokens)

| Tokens | Price | Prompt over 272K |
|---|---|---|
| Input | $1.15 | $2.30 |
| Cached input | $0.115 | $0.23 |
| Cache write (5 min) | $1.15 | $2.30 |
| Cache write (1 hour) | $2.30 | $4.60 |
| Output | $6.90 | $10.35 |

A request whose prompt is over 272K tokens is billed at $2.30 input and $10.35 output per million tokens.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.5","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-5.6 Luna API

> GPT-5.6 Luna by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $0.20 in / $1.20 out per 1M tokens.

- Model id: `gpt-5.6-luna`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 922K tokens
- Max output: 128K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-5.6-luna

### Price (USD per million tokens)

| Tokens | Price | Prompt over 272K |
|---|---|---|
| Input | $0.20 | $0.40 |
| Cached input | $0.02 | $0.04 |
| Cache write (5 min) | $0.25 | $0.50 |
| Cache write (1 hour) | $0.50 | $1.00 |
| Output | $1.20 | $1.80 |

A request whose prompt is over 272K tokens is billed at $0.40 input and $1.80 output per million tokens.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.6-luna","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-5.6 Sol API

> The GPT-5.6 flagship for complex professional work. By OpenAI, on Revana at $0.92 in / $4.60 out per 1M tokens.

- Model id: `gpt-5.6-sol`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1.05M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/gpt-5.6-sol

### Price (USD per million tokens)

| Tokens | Price | Prompt over 272K |
|---|---|---|
| Input | $0.92 | $1.84 |
| Cached input | $0.092 | $0.184 |
| Cache write (5 min) | $1.15 | $2.30 |
| Cache write (1 hour) | $1.84 | $3.68 |
| Output | $4.60 | $6.90 |

A request whose prompt is over 272K tokens is billed at $1.84 input and $6.90 output per million tokens.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.6-sol","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-5.6 Terra API

> The GPT-5.6 model that balances intelligence and cost. By OpenAI, on Revana at $0.46 in / $2.76 out per 1M tokens.

- Model id: `gpt-5.6-terra`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1.05M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/gpt-5.6-terra

### Price (USD per million tokens)

| Tokens | Price | Prompt over 272K |
|---|---|---|
| Input | $0.46 | $0.92 |
| Cached input | $0.046 | $0.092 |
| Cache write (5 min) | $0.575 | $1.15 |
| Cache write (1 hour) | $0.92 | $1.84 |
| Output | $2.76 | $4.14 |

A request whose prompt is over 272K tokens is billed at $0.92 input and $4.14 output per million tokens.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.6-terra","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-6 Astra API

> OpenAI's most capable model, built for the hardest end-to-end work. By OpenAI, on Revana at $2.30 in / $11.50 out per 1M tokens.

- Model id: `gpt-6-astra`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1.05M tokens
- Max output: 128K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-6-astra

### Price (USD per million tokens)

| Tokens | Price | Prompt over 272K |
|---|---|---|
| Input | $2.30 | $4.60 |
| Cached input | $0.23 | $0.46 |
| Cache write (5 min) | $2.875 | $5.75 |
| Cache write (1 hour) | $4.60 | $9.20 |
| Output | $11.50 | $17.25 |

A request whose prompt is over 272K tokens is billed at $4.60 input and $17.25 output per million tokens.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-6-astra","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-6 Luna API

> GPT-6 Luna by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $0.10 in / $0.50 out per 1M tokens.

- Model id: `gpt-6-luna`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 922K tokens
- Max output: 13K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-6-luna

### Price (USD per million tokens)

| Tokens | Price | Prompt over 272K |
|---|---|---|
| Input | $0.10 | $0.20 |
| Cached input | $0.01 | $0.02 |
| Cache write (5 min) | $0.125 | $0.25 |
| Cache write (1 hour) | $0.25 | $0.50 |
| Output | $0.50 | $0.75 |

A request whose prompt is over 272K tokens is billed at $0.20 input and $0.75 output per million tokens.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-6-luna","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-6 Sol API

> The GPT-6 model for complex coding and agentic workflows. By OpenAI, on Revana at $0.46 in / $2.30 out per 1M tokens.

- Model id: `gpt-6-sol`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1.05M tokens
- Max output: 128K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-6-sol

### Price (USD per million tokens)

| Tokens | Price | Prompt over 272K |
|---|---|---|
| Input | $0.46 | $0.92 |
| Cached input | $0.046 | $0.092 |
| Cache write (5 min) | $0.575 | $1.15 |
| Cache write (1 hour) | $0.92 | $1.84 |
| Output | $2.30 | $3.45 |

A request whose prompt is over 272K tokens is billed at $0.92 input and $3.45 output per million tokens.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-6-sol","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-6.1 Sol API

> Near-Astra performance for complex coding, computer use and professional work. By OpenAI, on Revana at $0.46 in / $2.30 out per 1M tokens.

- Model id: `gpt-6.1-sol`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1.05M tokens
- Max output: 128K tokens
- Routes: 1
- Page: https://revana.si/llm/gpt-6.1-sol

### Price (USD per million tokens)

| Tokens | Price | Prompt over 272K |
|---|---|---|
| Input | $0.46 | $0.92 |
| Cached input | $0.023 | $0.046 |
| Cache write (5 min) | $0.575 | $1.15 |
| Cache write (1 hour) | $0.92 | $1.84 |
| Output | $2.30 | $3.45 |

A request whose prompt is over 272K tokens is billed at $0.92 input and $3.45 output per million tokens.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-6.1-sol","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GPT-6.1 Terra API

> A fast, efficient model for visual reasoning, precise computer use and clear, natural writing. By OpenAI, on Revana at $0.23 in / $1.15 out per 1M tokens.

- Model id: `gpt-6.1-terra`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1.05M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/gpt-6.1-terra

### Price (USD per million tokens)

| Tokens | Price | Prompt over 272K |
|---|---|---|
| Input | $0.23 | $0.46 |
| Cached input | $0.0115 | $0.023 |
| Cache write (5 min) | $0.2875 | $0.575 |
| Cache write (1 hour) | $0.46 | $0.92 |
| Output | $1.15 | $1.725 |

A request whose prompt is over 272K tokens is billed at $0.46 input and $1.725 output per million tokens.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-6.1-terra","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## o3 API

> o3 by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $2.00 in / $8.00 out per 1M tokens.

- Model id: `o3`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/o3

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $2.00 |
| Cached input | $2.00 |
| Cache write (5 min) | $2.00 |
| Cache write (1 hour) | $2.00 |
| Output | $8.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"o3","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## o3-mini API

> o3-mini by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $1.10 in / $4.40 out per 1M tokens.

- Model id: `o3-mini`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/o3-mini

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.10 |
| Cached input | $0.55 |
| Cache write (5 min) | $1.10 |
| Cache write (1 hour) | $1.10 |
| Output | $4.40 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"o3-mini","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## o3-mini (high) API

> o3-mini (high) by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $1.10 in / $4.40 out per 1M tokens.

- Model id: `o3-mini-high`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/o3-mini-high

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.10 |
| Cached input | $1.10 |
| Cache write (5 min) | $1.10 |
| Cache write (1 hour) | $1.10 |
| Output | $4.40 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"o3-mini-high","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## o3-mini (low) API

> o3-mini (low) by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $1.10 in / $4.40 out per 1M tokens.

- Model id: `o3-mini-low`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/o3-mini-low

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.10 |
| Cached input | $1.10 |
| Cache write (5 min) | $1.10 |
| Cache write (1 hour) | $1.10 |
| Output | $4.40 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"o3-mini-low","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## o3-mini (medium) API

> o3-mini (medium) by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $1.10 in / $4.40 out per 1M tokens.

- Model id: `o3-mini-medium`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/o3-mini-medium

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.10 |
| Cached input | $1.10 |
| Cache write (5 min) | $1.10 |
| Cache write (1 hour) | $1.10 |
| Output | $4.40 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"o3-mini-medium","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## o4-mini API

> o4-mini by OpenAI, through an OpenAI-compatible chat completions API. By OpenAI, on Revana at $1.10 in / $4.40 out per 1M tokens.

- Model id: `o4-mini`
- Kind: text (chat completions)
- Maker: OpenAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 128K tokens
- Max output: 16K tokens
- Routes: 1
- Page: https://revana.si/llm/o4-mini

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.10 |
| Cached input | $0.275 |
| Cache write (5 min) | $1.10 |
| Cache write (1 hour) | $1.10 |
| Output | $4.40 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"o4-mini","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Qwen3.6-Flash API

> Qwen3.6-Flash by Qwen, through an OpenAI-compatible chat completions API. By Qwen, on Revana at $0.1714 in / $1.0284 out per 1M tokens.

- Model id: `qwen3.6-flash`
- Kind: text (chat completions)
- Maker: Qwen
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/qwen3.6-flash

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.1714 |
| Cached input | $0.1714 |
| Cache write (5 min) | $0.1714 |
| Cache write (1 hour) | $0.1714 |
| Output | $1.0284 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen3.6-flash","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Qwen3.6-Plus API

> Qwen3.6-Plus by Qwen, through an OpenAI-compatible chat completions API. By Qwen, on Revana at $0.2858 in / $1.7148 out per 1M tokens.

- Model id: `qwen3.6-plus`
- Kind: text (chat completions)
- Maker: Qwen
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/qwen3.6-plus

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.2858 |
| Cached input | $0.2858 |
| Cache write (5 min) | $0.2858 |
| Cache write (1 hour) | $0.2858 |
| Output | $1.7148 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen3.6-plus","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Qwen3.7-Flash API

> Qwen3.7-Flash by Qwen, through an OpenAI-compatible chat completions API. By Qwen, on Revana at $0.0286 in / $0.1143 out per 1M tokens.

- Model id: `qwen3.7-flash`
- Kind: text (chat completions)
- Maker: Qwen
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 984K tokens
- Max output: 131K tokens
- Routes: 1
- Page: https://revana.si/llm/qwen3.7-flash

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.0286 |
| Cached input | $0.0057 |
| Cache write (5 min) | $0.0357 |
| Cache write (1 hour) | $0.0357 |
| Output | $0.1143 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen3.7-flash","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Qwen3.7-Max API

> Qwen3.7-Max by Qwen, through an OpenAI-compatible chat completions API. By Qwen, on Revana at $1.7142 in / $5.1426 out per 1M tokens.

- Model id: `qwen3.7-max`
- Kind: text (chat completions)
- Maker: Qwen
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/qwen3.7-max

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.7142 |
| Cached input | $1.7142 |
| Cache write (5 min) | $1.7142 |
| Cache write (1 hour) | $1.7142 |
| Output | $5.1426 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen3.7-max","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Qwen3.7-Plus API

> Qwen3.7-Plus by Qwen, through an OpenAI-compatible chat completions API. By Qwen, on Revana at $0.2858 in / $1.1432 out per 1M tokens.

- Model id: `qwen3.7-plus`
- Kind: text (chat completions)
- Maker: Qwen
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/qwen3.7-plus

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.2858 |
| Cached input | $0.2858 |
| Cache write (5 min) | $0.2858 |
| Cache write (1 hour) | $0.2858 |
| Output | $1.1432 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen3.7-plus","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Qwen3.8 2.4T-A95B API

> Qwen3.8 2.4T-A95B by Qwen, through an OpenAI-compatible chat completions API. By Qwen, on Revana at $1.7143 in / $5.1429 out per 1M tokens.

- Model id: `qwen3.8-2.4t-a95b`
- Kind: text (chat completions)
- Maker: Qwen
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 984K tokens
- Max output: 131K tokens
- Routes: 1
- Page: https://revana.si/llm/qwen3.8-2.4t-a95b

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.7143 |
| Cached input | $0.2143 |
| Cache write (5 min) | $2.1429 |
| Cache write (1 hour) | $2.1429 |
| Output | $5.1429 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen3.8-2.4t-a95b","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Qwen3.8 27B API

> Qwen3.8 27B by Qwen, through an OpenAI-compatible chat completions API. By Qwen, on Revana at $0.4286 in / $1.7143 out per 1M tokens.

- Model id: `qwen3.8-27b`
- Kind: text (chat completions)
- Maker: Qwen
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 984K tokens
- Max output: 131K tokens
- Routes: 1
- Page: https://revana.si/llm/qwen3.8-27b

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.4286 |
| Cached input | $0.0857 |
| Cache write (5 min) | $0.5357 |
| Cache write (1 hour) | $0.5357 |
| Output | $1.7143 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen3.8-27b","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Qwen3.8-Flash API

> Qwen3.8-Flash by Qwen, through an OpenAI-compatible chat completions API. By Qwen, on Revana at $0.1143 in / $0.3857 out per 1M tokens.

- Model id: `qwen3.8-flash`
- Kind: text (chat completions)
- Maker: Qwen
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 984K tokens
- Max output: 131K tokens
- Routes: 1
- Page: https://revana.si/llm/qwen3.8-flash

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.1143 |
| Cached input | $0.0143 |
| Cache write (5 min) | $0.1786 |
| Cache write (1 hour) | $0.1786 |
| Output | $0.3857 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen3.8-flash","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Qwen3.8-Max API

> Alibaba's flagship Qwen model for complex reasoning, coding and agent work. By Qwen, on Revana at $1.20 in / $3.60 out per 1M tokens.

- Model id: `qwen3.8-max`
- Kind: text (chat completions)
- Maker: Qwen
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/qwen3.8-max

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.20 |
| Cached input | $0.15 |
| Cache write (5 min) | $1.50 |
| Cache write (1 hour) | $2.40 |
| Output | $3.60 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen3.8-max","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Step 3.7 Flash API

> Step 3.7 Flash by StepFun, through an OpenAI-compatible chat completions API. By StepFun, on Revana at $0.1928 in / $1.1568 out per 1M tokens.

- Model id: `step-3.7-flash`
- Kind: text (chat completions)
- Maker: StepFun
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/step-3.7-flash

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.1928 |
| Cached input | $0.1928 |
| Cache write (5 min) | $0.1928 |
| Cache write (1 hour) | $0.1928 |
| Output | $1.1568 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"step-3.7-flash","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Grok 4.3 API

> Grok 4.3 by xAI, through an OpenAI-compatible chat completions API. By xAI, on Revana at $1.25 in / $2.50 out per 1M tokens.

- Model id: `grok-4.3`
- Kind: text (chat completions)
- Maker: xAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/grok-4.3

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.25 |
| Cached input | $1.25 |
| Cache write (5 min) | $1.25 |
| Cache write (1 hour) | $1.25 |
| Output | $2.50 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"grok-4.3","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Grok 4.5 API

> xAI's coding model for agentic software development, engineering and workflow tasks. By xAI, on Revana at $0.60 in / $1.80 out per 1M tokens.

- Model id: `grok-4.5`
- Kind: text (chat completions)
- Maker: xAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 500K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/grok-4.5

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.60 |
| Cached input | $0.09 |
| Cache write (5 min) | $0.60 |
| Cache write (1 hour) | $1.20 |
| Output | $1.80 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"grok-4.5","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Grok 4.6 API

> Grok 4.6 by xAI, through an OpenAI-compatible chat completions API. By xAI, on Revana at $2.00 in / $6.00 out per 1M tokens.

- Model id: `grok-4.6`
- Kind: text (chat completions)
- Maker: xAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 500K tokens
- Max output: 13K tokens
- Routes: 1
- Page: https://revana.si/llm/grok-4.6

### Price (USD per million tokens)

| Tokens | Price | Prompt over 200K |
|---|---|---|
| Input | $2.00 | $4.00 |
| Cached input | $0.50 | $1.00 |
| Cache write (5 min) | $2.00 | $4.00 |
| Cache write (1 hour) | $2.00 | $4.00 |
| Output | $6.00 | $12.00 |

A request whose prompt is over 200K tokens is billed at $4.00 input and $12.00 output per million tokens.
Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"grok-4.6","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Grok 4.7 API

> xAI's most capable model for coding and knowledge work. By xAI, on Revana at $0.60 in / $1.80 out per 1M tokens.

- Model id: `grok-4.7`
- Kind: text (chat completions)
- Maker: xAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 500K tokens
- Max output: 13K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/grok-4.7

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.60 |
| Cached input | $0.15 |
| Cache write (5 min) | $0.60 |
| Cache write (1 hour) | $1.20 |
| Output | $1.80 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"grok-4.7","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Grok 4.20 (non-reasoning) API

> Grok 4.20 (non-reasoning) by xAI, through an OpenAI-compatible chat completions API. By xAI, on Revana at $1.25 in / $2.50 out per 1M tokens.

- Model id: `grok-4.20-0309-non-reasoning`
- Kind: text (chat completions)
- Maker: xAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/grok-4.20-0309-non-reasoning

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.25 |
| Cached input | $1.25 |
| Cache write (5 min) | $1.25 |
| Cache write (1 hour) | $1.25 |
| Output | $2.50 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"grok-4.20-0309-non-reasoning","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Grok 4.20 (reasoning) API

> Grok 4.20 (reasoning) by xAI, through an OpenAI-compatible chat completions API. By xAI, on Revana at $1.25 in / $2.50 out per 1M tokens.

- Model id: `grok-4.20-0309-reasoning`
- Kind: text (chat completions)
- Maker: xAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/grok-4.20-0309-reasoning

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.25 |
| Cached input | $1.25 |
| Cache write (5 min) | $1.25 |
| Cache write (1 hour) | $1.25 |
| Output | $2.50 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"grok-4.20-0309-reasoning","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Grok 4.20 Multi-Agent API

> Grok 4.20 Multi-Agent by xAI, through an OpenAI-compatible chat completions API. By xAI, on Revana at $1.25 in / $2.50 out per 1M tokens.

- Model id: `grok-4.20-multi-agent-0309`
- Kind: text (chat completions)
- Maker: xAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/grok-4.20-multi-agent-0309

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.25 |
| Cached input | $1.25 |
| Cache write (5 min) | $1.25 |
| Cache write (1 hour) | $1.25 |
| Output | $2.50 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"grok-4.20-multi-agent-0309","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## Grok Build 0.1 API

> Grok Build 0.1 by xAI, through an OpenAI-compatible chat completions API. By xAI, on Revana at $1.00 in / $2.00 out per 1M tokens.

- Model id: `grok-build-0.1`
- Kind: text (chat completions)
- Maker: xAI
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/grok-build-0.1

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.00 |
| Cached input | $1.00 |
| Cache write (5 min) | $1.00 |
| Cache write (1 hour) | $1.00 |
| Output | $2.00 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"grok-build-0.1","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GLM-5 API

> GLM-5 by Zhipu GLM, through an OpenAI-compatible chat completions API. By Zhipu GLM, on Revana at $1.00 in / $1.60 out per 1M tokens.

- Model id: `glm-5`
- Kind: text (chat completions)
- Maker: Zhipu GLM
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Routes: 1
- Page: https://revana.si/llm/glm-5

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $1.00 |
| Cached input | $1.00 |
| Cache write (5 min) | $1.00 |
| Cache write (1 hour) | $1.00 |
| Output | $1.60 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"glm-5","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GLM-5.1 API

> Z.ai's earlier general-purpose GLM for chat and coding. By Zhipu GLM, on Revana at $0.84 in / $2.64 out per 1M tokens.

- Model id: `glm-5.1`
- Kind: text (chat completions)
- Maker: Zhipu GLM
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 200K tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/glm-5.1

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.84 |
| Cached input | $0.156 |
| Cache write (5 min) | $0.84 |
| Cache write (1 hour) | $1.68 |
| Output | $2.64 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"glm-5.1","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GLM-5.2 API

> Z.ai's general-purpose GLM for reasoning and coding. By Zhipu GLM, on Revana at $0.84 in / $2.64 out per 1M tokens.

- Model id: `glm-5.2`
- Kind: text (chat completions)
- Maker: Zhipu GLM
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/glm-5.2

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.84 |
| Cached input | $0.156 |
| Cache write (5 min) | $0.84 |
| Cache write (1 hour) | $1.68 |
| Output | $2.64 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"glm-5.2","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GLM-5.3 API

> Z.ai's flagship for complex software engineering and long-horizon agent work. By Zhipu GLM, on Revana at $0.84 in / $2.64 out per 1M tokens.

- Model id: `glm-5.3`
- Kind: text (chat completions)
- Maker: Zhipu GLM
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/glm-5.3

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.84 |
| Cached input | $0.156 |
| Cache write (5 min) | $0.84 |
| Cache write (1 hour) | $1.68 |
| Output | $2.64 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"glm-5.3","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GLM-5.3 Flash API

> The lowest-cost GLM-5, and the first in the series to read images. By Zhipu GLM, on Revana at $0.09 in / $0.30 out per 1M tokens.

- Model id: `glm-5.3-flash`
- Kind: text (chat completions)
- Maker: Zhipu GLM
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 128K tokens
- Routes: 2 (starts on the cheapest, falls back automatically)
- Page: https://revana.si/llm/glm-5.3-flash

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.09 |
| Cached input | $0.018 |
| Cache write (5 min) | $0.09 |
| Cache write (1 hour) | $0.18 |
| Output | $0.30 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"glm-5.3-flash","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat

## GLM-5.3 FlashX API

> GLM-5.3 Flash at about 200 tokens a second. By Zhipu GLM, on Revana at $0.222 in / $0.75 out per 1M tokens.

- Model id: `glm-5.3-flashx`
- Kind: text (chat completions)
- Maker: Zhipu GLM
- Endpoint: `POST https://api.revana.si/v1/chat/completions` (OpenAI-compatible, streaming supported)
- Context window: 1M tokens
- Max output: 128K tokens
- Routes: 1
- Page: https://revana.si/llm/glm-5.3-flashx

### Price (USD per million tokens)

| Tokens | Price |
|---|---|
| Input | $0.222 |
| Cached input | $0.045 |
| Cache write (5 min) | $0.222 |
| Cache write (1 hour) | $0.444 |
| Output | $0.75 |

Reasoning tokens are billed as output. Each answer's `usage.cost_usd` is the amount taken from your balance; a request with no answer costs nothing.

### Request

```bash
curl https://api.revana.si/v1/chat/completions \
  -H "Authorization: Bearer $REVANA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"glm-5.3-flashx","messages":[{"role":"user","content":"Write a haiku about routing."}]}'
```

Or any OpenAI SDK with `base_url="https://api.revana.si/v1"`. Docs: https://revana.si/chat