Domestic Video Model Generation
Domestic Video Model Generation
Submit domestic video tasks using POST /v1/videos.
POST
Domestic Video Model Generation
Documentation Index
Fetch the complete documentation index at: https://doc.deepwl.cn/llms.txt
Use this file to discover all available pages before exploring further.
Domestic Video Model Generation
Method and Path
Request Example
Fields
Model name. You can pass either a base model directly or a bundled billing model.
Prompt.
Duration. The top-level
seconds has the highest priority.Duration compatibility field.
Quick size field, such as
720x1280 or 1280x720.Single reference image or first-frame image.
Multiple reference images.
Reference image input, compatible with URL, base64, data URI, and file upload.
Extended parameters. Commonly used
metadata.output_config, scene_type, motion_level, offpeak, video_url, file_infos, ext_info.Scenarios
| Scenario | Key Fields |
|---|---|
| Text-to-video | prompt + seconds |
| Image-to-video | image / images / input_reference |
| First and last frame | image + metadata.last_frame_url |
| Motion control | metadata.scene_type=motion_control + video reference |
| Digital human | metadata.scene_type=avatar_i2v |
| Lip sync | metadata.scene_type=lip_sync |
| Template effects | metadata.scene_type=template_effect |
Rules
seconds/durationcan be placed at the top level or inmetadata- The top-level duration has the highest priority
scene_type=motion_controlmust provide a video referencescene_type=lip_syncis counted as 5 seconds if it is less than 5 secondsFileInfoscan contain up to 3 items
metadata.output_config
Common fields:
resolutionaspect_ratioaudio_generationdurationperson_generationinput_compliance_checkoutput_compliance_checkenhance_switchframe_interpolatelogo_add