Image Generation

POST

/kling/v1/images/generations

Create an image generation task from a text prompt and optional reference image.

Authorizations

bearer

Type

HTTP (bearer)

Request Body

application/json

object

Model name. Enum: kling-v1, kling-v1-5, kling-v2, kling-v2-new, kling-v2-1, kling-v3

Positive text prompt (max 2500 characters)

Negative text prompt (max 2500 characters)

Reference image as Base64 or image URL (must be accessible)

Image reference type.
Enum: subject (character traits), face (facial likeness).
When using face, the uploaded image must contain exactly one face.
Required when using kling-v1-5 with a non-empty image parameter.

Reference strength for the uploaded image during generation.
Range: [0, 1]; higher values mean stronger reference.

Facial reference strength (similarity of facial features to the reference).
Only applies when image_reference is subject.
Range: [0, 1]; higher values mean stronger reference.

Output image resolution.
Enum: 1k, 2k
1k: 1K standard
2k: 2K HD

Number of images to generate.
Range: [1, 9]

Output aspect ratio (width:height).
Enum: 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, 2:3, 21:9

Callback URL when the task completes

Responses

OK

Content-Type

application/json

object

Task ID

Task status