Create Video Generation Task

Submit a video generation task, supporting text-to-video and image-to-video.

Returns the task ID, which can be queried via the GET interface.

General Video Tasks

This entry supports most text-to-video and image-to-video channels. After creation, use GET /v1/video/generations/{task_id} to query; general fields prioritize using model, prompt, image/images, input_reference, seconds, size, with channel-specific parameters placed in metadata.

curl "$BASE_URL/v1/video/generations" \
  -H "Authorization: Bearer $NODEKEY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "doubao-seedance-2-0-260128",
    "prompt": "A handheld camera moves through a coffee shop in the morning, sunlight sweeping across the table from the window",
    "seconds": "5",
    "size": "1280x720",
    "metadata": {"ratio": "16:9", "resolution": "720p", "generate_audio": false}
  }'

curl "$BASE_URL/v1/video/generations/task_xxx" \
  -H "Authorization: Bearer $NODEKEY_API_KEY"

Non-standard interfaces retain channel formats: Keling uses /kling/v1/videos/text2video, /kling/v1/videos/image2video; Jimeng uses /jimeng/?Action=.... Seedance's metadata.content can pass image_url, video_url, audio_url as reference content.

curl -X POST "https://nodekey.xinghanyun.cn/v1/video/generations" \  -H "Content-Type: application/json" \  -d '{    "model": "kling-v1",    "prompt": "宇航员站起身走了",    "image": "https://example.com/image.jpg",    "duration": 5,    "width": 1280,    "height": 720,    "fps": 30,    "seed": 20231234,    "n": 1,    "response_format": "url",    "user": "user-1234",    "metadata": {}  }'

{
  "task_id": "abcd1234efgh",
  "status": "queued"
}

{
  "error": {
    "message": "string",
    "type": "string",
    "param": "string",
    "code": "string"
  }
}