Kling video
The Kling video family uses the OpenAI-compatible /v1/videos API and supports text-to-video, image-to-video, motion control, digital avatar, lip sync, and other generation modes.
OpenAI format
Submit with POST /v1/videos and query with GET /v1/videos/{task_id}.
Video generation
`POST /v1/videos` — submit a Kling video task
View docs →Task status query
`GET /v1/videos/{task_id}` — query task status and results
View docs →Supported models
| Model | Description | Duration rules |
|---|---|---|
| Kling 3.0 all-in-one | 5–15 seconds | |
| Kling 2.6 | 5–10 seconds | |
| Kling 2.5 | 5–10 seconds |
Scenario types
| Scenario type | Description |
|---|---|
| Motion control | |
| Digital avatar generation | |
| Lip sync | |
| Template effects |
Basic parameters
| Field | Type | Description |
|---|---|---|
| string | Model name | |
| string | Prompt | |
| string | Video duration | |
| string | Video size | |
| string | Reference image |
Endpoint summary
| Capability | Method | Path |
|---|---|---|
| Video generation | POST | |
| Task query | GET |