Generate video
Submits a video generation request. Returns 202 Accepted immediately with a job_uid; poll GET /v1/jobs/{job_uid} for status and the final video URL.
Before submitting: call GET /v1/videos/models to retrieve the valid parameter values (duration, resolution, mode, aspect_ratio, etc.) for your chosen model. Passing an unsupported value returns a 422 validation error.
Billing: the cost is deducted from your API wallet when the job is accepted. If video generation fails on the provider side the charge is automatically refunded. Requires scope videos:write.
Authorizations
API key from Developer dashboard (Bearer token).
Body
Model identifier. See GET /v1/videos/models for available values.
"kling-v2-1"
Generation mode. text_to_video requires prompt. image_to_video requires image_url; prompt is optional motion guidance.
text_to_video, image_to_video "text_to_video"
Text description of the video. Required for text_to_video. Optional for image_to_video. Max length varies per model — see prompt_max_length in /v1/videos/models.
"A lone astronaut walks across a crimson desert at sunset"
Source image URL for image_to_video. Must be a publicly accessible JPEG or PNG.
"https://example.com/my-image.jpg"
Video length in seconds. Valid values vary per model — see duration in /v1/videos/models. Defaults to the shortest valid duration.
10
Output resolution (e.g. 720p, 1080p). Valid values vary per model — see resolution in /v1/videos/models.
"1080p"
Generation quality mode (e.g. Standard, Professional, Fast). Valid values vary per model — see mode in /v1/videos/models.
"Professional"
Output aspect ratio (e.g. 16:9, 9:16, 1:1). Valid values vary per model — see aspect_ratio in /v1/videos/models.
"16:9"
Visual style preset (e.g. Anime, Cyberpunk). Only supported by some models — see style in /v1/videos/models.
"Anime"
Camera/subject movement amplitude (Auto, Small, Medium, Large). Only supported by some models — see movement in /v1/videos/models.
"Medium"
Elements to suppress in the output. Only supported by some models — see supports_negative_prompt in /v1/videos/models.
"blurry, low quality, watermark"
Reproducibility seed. Only supported by some models — see supports_seed in /v1/videos/models.
42
Generate ambient audio alongside the video. Only supported by some models — see supports_generate_audio in /v1/videos/models.
false
Lock the camera position (no camera movement). Only supported by some models — see supports_camera_fixed in /v1/videos/models.
false

