API Reference
Interface Overview
| Interface | Method | Description |
|---|---|---|
| /v1/chat/completions | POST | Create chat completion |
| /v1/models | GET | Get model list |
| /v1/usage | GET | Get usage statistics |
Data Models
Message Object
json
{
"role": "user|assistant|system|tool",
"content": "Message content",
"name": "Sender name",
"tool_calls": [...],
"tool_call_id": "tool_call_id"
}Choice Object
json
{
"index": 0,
"message": {...},
"finish_reason": "stop|length|tool_calls|content_filter"
}Usage Object
json
{
"prompt_tokens": 100,
"completion_tokens": 50,
"total_tokens": 150
}Kling Video Models
Supported Models
| Model ID | Description | Recommended Endpoint |
|---|---|---|
kling-v1 | Kling V1.0 Base | text2video / image2video |
kling-v1-5 | Kling V1.5 | text2video / image2video |
kling-v1-6 | Kling V1.6 | text2video / image2video |
kling-v2-1 | Kling V2.1 | text2video / image2video |
kling-v2-5-turbo | Kling V2.5 Turbo | text2video / image2video |
kling-v2-6 | Kling V2.6 | text2video / image2video |
kling-video-o1 | Kling Omni V1 | omni-video |
kling-v3 | Kling V3.0 Text/Image to Video | text2video / image2video |
kling-v3-omni | Kling V3.0 Omni Video (Latest) | omni-video |
kling-v2-1-master | Kling V2.1 Master | text2video / image2video |
kling-v2-master | Kling V2 Master | text2video / image2video |
Kling V3 New Features
kling-v3-omni provides the following new capabilities via the Omni Video endpoint:
- Extended Duration: 3-15 seconds (O1 supports 3-10 seconds)
- Multi-shot Narrative: Generate multiple consecutive shots in a single request
- Video Editing: Edit existing videos with text instructions via
refer_type: "base" - Native Audio: Generate synchronized audio (with multilingual lip sync)