Create Veo 3.1 Video

POST 
/v1/videos

Create an asynchronous Veo 3.1 video generation task with the OpenAI-compatible Video API. Veo 3.1 is suitable for high-fidelity video generation, including text-to-video, image-referenced video generation, and synchronized audio when supported by the configured channel and upstream model.

Flow

1. Call POST /v1/videos to submit the video generation task.
2. Read id or task_id from the response.
3. Poll GET /v1/videos/{task_id} for task status and results.

Example

curl --location 'https://newapi.xxd.fans/v1/videos' \
  --header 'Content-Type: application/json' \
  --header 'Authorization: Bearer YOUR_API_KEY' \
  --data '{
    "model": "veo-3.1",
    "prompt": "A serene mountain landscape at sunset with clouds drifting by"
  }'

If the configured upstream channel uses the official Google Gemini or Vertex Veo model name, map veo-3.1 to the actual upstream model in the dashboard.

Request

Responses

200

Task submitted successfully

400

Invalid request parameters

401

Unauthorized

502

Upstream task submission failed

Response Example

{
  "id": "task_DCAvd1SdbNi3Gv6NAt5uoSS3tbyCuZ4m",
  "task_id": "task_DCAvd1SdbNi3Gv6NAt5uoSS3tbyCuZ4m",
  "object": "video",
  "model": "veo-3.1",
  "status": "queued",
  "progress": 0,
  "created_at": 1710000000
}

Notes

• Use veo-3.1 as the public model name. Configure model mapping if the upstream provider requires another model name.
• Video generation is asynchronous. A successful submission does not mean the video has already finished.
• If the upstream provider returns 201 Created or 202 Accepted, the gateway version must accept 2xx responses as successful task submissions.