Cancel a queued fal request (best-effort). Provide the Cancel URL returned by Submit Request, or the Model Endpoint Id together with the Request Id. A request already in progress may not stop.
Generate images from a text prompt using a fal text-to-image model (default fal-ai/flux/schnell). Runs synchronously and returns the generated image URLs.
Download a fal CDN media URL into file storage. fal outputs expire per account settings, so use this to persist generated media.
Search fal's model catalogue. Filter by query, category and status. Uses cursor-based pagination (no offset).
Fetch the result of a completed fal queue request. Provide the Response URL returned by Submit Request, or the Model Endpoint Id together with the Request Id.
Get the status of a queued fal request. Provide the Status URL returned by Submit Request, or the Model Endpoint Id together with the Request Id.
Perform an arbitrary authorized API call to fal. Pick one of fal's base URLs, then provide a path (or full URL), method, headers, query parameters and body.
Run any fal.ai model by its endpoint id. Set Wait For Result on for a synchronous call (bounded to 60s) or off to submit the job to the async queue and return immediately.
Submit a model inference job to the fal async queue and return immediately with the request handles. Pair the optional Webhook URL with a Generation Completed trigger, or poll with Get Request Status / Get Request Result.
Upload a file to the fal CDN and return a public URL that can be passed as a model input (for image, audio or video inputs).