Ảnh sang video

POST

/kling/v1/videos/image2video

Tạo tác vụ tạo video từ ảnh tham chiếu.

Authorizations

bearer

Type

HTTP (bearer)

Request Body

application/json

object

Model name. Enum: kling-v1, kling-v1-5, kling-v1-6, kling-v2-master, kling-v2-1, kling-v2-1-master, kling-v2-5-turbo, kling-v2-6, kling-v3

Reference image. Base64-encoded image or image URL (must be accessible).

Reference image — end-frame control. Base64-encoded image or image URL (must be accessible).

Positive text prompt, up to 2500 characters.
Use <<<voice_1>>> to specify a voice; the index matches the order in voice_list.
At most 2 voices per task; when a voice is specified, sound must be on. Keep syntax simple, e.g.:
A man <<<vocie_1>>> says: "Hello"
When voice_list is set and the prompt references a voice ID, billing uses the "with specified voice" rate.

Negative text prompt

object[]

Supported on V2.6 and later. List of voices used for video generation.
At most 2 voices per task.
When voice_list is set and the prompt references a voice ID, billing uses the "with specified voice" rate.

Supported on V2.6 and later. Whether to generate audio with the video.
Enum: on, off

Not supported on kling-v2.x models. Controls freedom; higher values mean less freedom and stronger adherence to the prompt.

Video generation mode.
Enum: std, pro
std: Standard mode — balanced quality and cost.
pro: Pro mode — higher quality output.

Static brush mask region (mask image from motion brush painting).

Whether to generate a multi-shot video.
If true, prompt is ignored.
If false, shot_type and multi_prompt are ignored.

Shot segmentation mode.
Enum: customize
Required when multi_shot is true.

object[]

Per-shot details (prompt, duration, etc.).
● Define shot index, prompt, and duration via index, prompt, and duration:
○ Up to 6 shots, minimum 1
○ Max 512 characters per shot content
○ Each shot duration ≤ total task duration and ≥ 1
○ Sum of shot durations equals total task duration

object[]

Dynamic brush configuration list

object