Skip to main content
Use Studio AI endpoints to generate images and videos with the same model catalog and credit rules used in-app. All routes are key-scoped and use the standard Partner API response envelope documented in Response Pattern.

Endpoints

Polling Flow

Image and video generation are both asynchronous.
  1. Call POST /studio/images/generate or POST /studio/videos/generate.
  2. Store data.image_id or data.video_id. Credits are charged at this point.
  3. Poll GET /studio/images/{imageId} or GET /studio/videos/{videoId} until data.status is succeeded or failed.
  4. Use data.output_url once status is succeeded.
If you lose the id, List Studio Creations shows recent results.

Input Contract

Studio generation endpoints use normalized partner input (snake_case) with optional raw_params for model-specific passthrough:
  • params: stable normalized keys (aspect_ratio, duration_seconds, etc.)
  • raw_params: optional advanced passthrough object
This keeps client payloads consistent while still allowing model-specific controls.