# i2v.ai > i2v.ai provides one API and a browser-based Studio for generating videos and creating or editing images with multiple models. REST, MCP, and Studio share an account balance in USD. Choose a model by its public API ID and check its document for supported inputs and published USD prices. Rates vary by model, configuration, and billing unit. Use an [API key](https://i2v.ai/me?section=api-keys) from your account. Keep it server-side. REST uses `https://api.i2v.ai` with `Authorization: Bearer YOUR_API_KEY`. For an image or video, send the exact model and inputs to `POST /api/v1/jobs/quote`. Review `maximum.amount` and confirm the USD budget before submitting. A quote creates no task and reserves no funds. Submit the unchanged model and inputs to `POST /api/v1/jobs/createTask`. Pass `maxCostUsd` as the approved USD decimal string, `expectedRevision` from the quote, and a stable `Idempotency-Key` header. Persist that key, payload, and the returned task ID for recovery. Poll `GET /api/v1/jobs/recordInfo?taskId=TASK_ID` on the same account. Back off while `status` is `waiting` or `generating`. Read `output[].url` on `success` and `error.code` and `error.message` on `fail`. If the submission response is lost, check `GET /api/v1/jobs/submission` with the original key and credentials. Follow `Retry-After` when provided. For MCP, connect to `https://i2v.ai/api/mcp` over Streamable HTTP. Start with read-only model discovery. Use `prepare_generation` to check parameters and quote before submitting. See the MCP guide for client setup and write authorization. ## When to use i2v.ai - [Turn a prompt into a video with Google Veo 3.1](https://i2v.ai/api/docs/markdown?path=models/google-veo3-1): Generate an 8-second clip with audio from a text prompt. API ID: `google-veo3.1`. - [Animate a still image with Wan 3.0 Prime](https://i2v.ai/api/docs/markdown?path=models/wan-3-0-pro): Use an image as the starting frame for a 2–30 second clip at 480p, 720p, or 1080p. API ID: `wan-3-0-pro`. - [Generate a video from reference media with Seedance 2.5](https://i2v.ai/api/docs/markdown?path=models/seedance2-5): Use reference images, video, or audio to guide a new clip. API ID: `seedance2-5`. - [Edit an existing video with Kling 3.0 Omni](https://i2v.ai/api/docs/markdown?path=models/official-kling-v3-omni): Apply a prompt to a 3–15 second source clip in video-to-video mode. API ID: `official-kling-v3-omni`. - [Generate an image from text with GPT Image 2](https://i2v.ai/api/docs/markdown?path=models/gpt-image-2): Choose 1K, 2K, or 4K output and check the allowed aspect ratios for that resolution. API ID: `gpt-image-2`. - [Edit an image with Nano Banana 2](https://i2v.ai/api/docs/markdown?path=models/nanobanana-2): Supply one or more reference images and describe the changes in a prompt. API ID: `nanobanana-2`. ## Get started - [REST quickstart](https://i2v.ai/api/docs/markdown): Quote the exact inputs, submit within the USD ceiling, and retrieve the result. - [API reference](https://i2v.ai/docs/api): Task and account HTTP contracts. - [OpenAPI](https://i2v.ai/openapi.json): Machine-readable public API contract. - [MCP guide](https://i2v.ai/api/docs/markdown?path=mcp): Connect an agent, check parameters and quote, then submit or query tasks. - [MCP setup](https://i2v.ai/mcp): Copy setup for your client and verify the connection with a read-only prompt. - [CLI and Agent Skills](https://i2v.ai/cli): Install i2v-models or i2v-integrate, download the CLI, or connect a local MCP client. - [Model API directory](https://i2v.ai/docs/models): Find public model IDs, input constraints, runnable examples, and per-configuration prices. - [Model catalog JSON](https://i2v.ai/api/v1/generation-models): Discover image and video model IDs, input schemas, and published prices programmatically. Use this endpoint for media models; `/api/v1/models` is the Chat Completions model list. - [SaaS and workflow integration](https://i2v.ai/api/docs/markdown?path=integration): Own tasks in your backend and verify delivery. - [Media quote, upload and recovery](https://i2v.ai/api/docs/markdown?path=media-workflow): Use signed upload tickets and recover lost submissions. - [Signed webhooks](https://i2v.ai/api/docs/markdown?path=webhooks): Verify and deduplicate completion events. - [Runnable API examples](https://github.com/i2v-ai/i2v-api-examples): Start an integration in Node.js, Python, or curl. - [Errors and retries](https://i2v.ai/api/docs/markdown?path=errors): Recover interrupted requests with the original idempotency key and handle rate limits. ## Check pricing and availability - [USD billing guide](https://i2v.ai/api/docs/markdown?path=billing): Model pricing, USD reservations, settlement, and confirmed failures. - [Pricing](https://i2v.ai/pricing): Public purchase options and pricing overview. - [Model status Markdown](https://i2v.ai/status.md): Current image and video model status and recent exceptions. - [Model status JSON](https://i2v.ai/api/status): Public status snapshot and daily history. - [Service status page](https://i2v.ai/status): Browse model availability in the browser. ## Optional - [Studio](https://i2v.ai/app): Generate images and videos in the browser. - [Full documentation and model index](https://i2v.ai/llms-full.txt): Browse all public models listed in the developer directory and the full set of guides.