MaiToken
Login/Register

Sora 2 Video Generation

Native OpenAI Videos protocol: submit a task, poll status, download the video

POST/v1/videos

Sora models use the native OpenAI Videos API, transparently forwarded. Available model names (e.g. sora-2) come from List Models.

Authorizations

AuthorizationstringRequired

Bearer Token authentication.

Authorization: Bearer YOUR_API_KEY

Submit a task

POST /v1/videos

Native OpenAI Videos body (fields per official OpenAI docs):

  • model (required), prompt (required)
  • seconds (optional, string): e.g. "4" / "8" / "12"
  • size (optional): e.g. "720x1280" / "1280x720"

Poll status

GET /v1/videos/{video_id}

Status values (native OpenAI): queued / in_progress / completed / failed. Poll every 5–10 seconds.

Download the video

GET /v1/videos/{video_id}/content

Once completed, download the binary MP4 from the content endpoint.

Notes

  • Query / download with the same API Key used for submission; task Ids are queryable for 7 days, then 404 task_not_found
  • For re-creation based on a completed video see Remix
  • Full field definitions follow the official OpenAI Videos documentation; no field rewriting