Documentation/OpenAI video tasks

OpenAI video tasks

Behavior

After creation, GET /v1/videos/{id}; when completed GET /v1/videos/{id}/content. Poll with backoff and the same caller. No universal cancel endpoint is provided.

Endpoint and authentication

POST /v1/videos

Use an XHub API key: Authorization: Bearer . Do not use upstream provider credentials as gateway keys.

Use the gateway root Base URL shown in the playground, without appending /v1; send requests to the full endpoint path above.

Request example

curl "$XHUB_BASE_URL/v1/videos" \
  -H "Authorization: Bearer $XHUB_API_KEY" \
  -F "model=YOUR_MODEL_NAME" -F "prompt=A lighthouse at sunrise." -F "seconds=8"

Request parameters

Field Type Description
model string · required Public model name; deployment must support this protocol.
prompt string · required Video description.
seconds / size / input_reference string / file Supported duration, resolution and optional reference.

Success response

{
  "id": "video-example",
  "status": "queued"
}
Field Type
id string
status string

Errors and troubleshooting

HTTP Meaning
400 Invalid request. Check JSON, parameters and query.
401 Missing or invalid XHub credential.
403 Permission denied. Check model scope or policy.
429 Limit exceeded. Inspect response details and Retry-After.

Search full documentation: budgets, routing, Docker…