API
OpenAPI llms.txt Create API key

AITOPIA REST API

Generate images, videos and audio from your own application with one API key.

The AITOPIA REST API lets your server run AITOPIA models, edit existing media and submit batches of generations over plain HTTPS. It is the same generation layer the AITOPIA app and the AITOPIA MCP server use: same models, same prices, same input validation and the same output checks.

  • Base URLs: live https://api-live.aitopia.ai, sandbox https://api-sandbox.aitopia.ai (test keys, simulated runs, nothing charged; see Sandbox). Every path starts with /v1, for example https://api-live.aitopia.ai/v1/models.
  • Auth: Authorization: Key <your key> (Authentication)
  • Format: JSON request and response bodies, snake_case field names
  • Billing: credits come from the AITOPIA account that owns the key (Billing & retention)

#How it works

Generation is asynchronous. A submit call answers at once with 202 Accepted and a request_id; you then poll the request or receive a signed webhook when it finishes.

Text
POST /v1/models/{model_id}   ->  202 { request_id, status: "queued", status_url, cancel_url, ... }
GET  /v1/requests/{id}       ->  200 { status: "running", ... }
GET  /v1/requests/{id}       ->  200 { status: "completed", output: { images: [{ url, content_type }] } }

Read Requests & lifecycle for the full picture, then Polling or Webhooks.

#What you can call

EndpointPurpose
GET /v1/modelsList and search the models you can run
GET /v1/models/{model_id}Input schema and price of one model
POST /v1/models/{model_id}Run a model
POST /v1/editEdit an existing image, video or audio file from a plain-language instruction
POST /v1/batchSubmit 1 to 12 independent generations as one request
GET /v1/requests/{request_id}Status and result of a request
POST /v1/requests/{request_id}/cancelAsk to cancel a request
POST /v1/uploadsHost a file (base64 or import from a URL)
POST /v1/uploads/linkOne-time upload URL for files up to 95 MB
GET /v1/account/balanceCredit balance of the key owner

Every endpoint is described in the Endpoints reference.

#Models

The catalogue changes often, so models are not listed in these docs. Call GET /v1/models (filter with type and q) to see what you can run, and GET /v1/models/{model_id} for a model's input fields and price. Model ids contain a slash, for example google/nano-banana-2 (image), minimax/h3-max or google/veo-3.1-lite (video), minimax/speech-2.8-hd (speech) and minimax/music-3 (music). Send the slash as it is or encoded as %2F (OpenAPI-generated clients encode it); both reach the same model. The examples in these docs send only prompt; real models take more fields (images, durations, resolutions), listed by GET /v1/models/{model_id}.

#Machine-readable files

  • openapi.json: OpenAPI 3.1 description of every endpoint. The API also serves it, without a key, at GET https://api-live.aitopia.ai/v1/openapi.json.
  • llms.txt: a compact index of these docs for AI agents and coding assistants.

#Next steps

AITOPIA REST API · Live https://api-live.aitopia.ai/v1 · Sandbox https://api-sandbox.aitopia.ai/v1