Kling

Call Kling through the OpenAI-style async Video API:

Action	Method	Path
Create	`POST`	`/v1/videos`
Get status	`GET`	`/v1/videos/{task_id}`
Download	`GET`	`/v1/videos/{task_id}/content`

Host: https://api.omniall.ai · Auth Authorization: Bearer sk-....

Models

Generation aliases (case-insensitive): kling-1.0, kling-1.5, kling-2.0, kling-2.1, kling-2.5, kling-3.0, kling-3.0-omni, kling-o1. Specialty: kling-identify-face (lip sync), kling-avatar-720p, kling-avatar-1080p. See the Model Square. Status responses may show the concrete billing tier.

Defaults & priority

Item	Rule
Duration	`seconds` > `duration` > metadata duration fields > 5s
Resolution	`output_config.resolution` > `metadata.resolution` > `size` > 720P
Audio	Default Disabled (mute)
Motion tier	Default std
T2V aspect	Default 16:9

Scenarios

Scenario	Key fields
Text-to-video	`prompt` + `seconds` + `output_config`
Single image	`image` + `prompt`
Multi-image	`images[]` (Omni)
First / last frame	`image` + `metadata.last_frame_url`
Motion control	`scene_type=motion_control` + public `video_url`
Avatar	`scene_type=avatar_i2v` or specialty model
Lip sync	`scene_type=lip_sync` or `kling-identify-face`
Reference video	`metadata.file_infos`
Multi-shot	`metadata.ext_info` (JSON string)

Motion control requires a publicly reachable reference video URL.