Kling 4.0 API

Zurelay charges by the second of video, sound included: $0.15 per second at 720p and $0.225 at 1080p. A clip costs its length times that rate, so a 5 second 720p clip is $0.75 and a 30 second 1080p clip is 30 times $0.225. First and last frames, reference images and the aspect ratio don't change the price.

When Zurelay accepts the job, its price is held from your balance, so videos in progress can never spend more than you have. It's charged only when the video is delivered. If a video fails, including when the content filter refuses the prompt or images, you pay nothing.

Send POST https://api.zurelay.com/v1/videos with your Zurelay key as a Bearer token and a JSON body: model set to kling-4.0, a prompt of up to 8,000 characters, and optionally seconds (default 5), resolution (720p by default, or 1080p) and aspect_ratio (default 16:9). You get a job back right away. Poll GET /v1/videos/{id} every few seconds until its status is completed, then download the MP4 from /v1/videos/{id}/content. Any HTTP client works; there's no SDK to install.

Any whole length from 3 to 30 seconds, made in one generation rather than stitched together. Longer clips take longer to make: a 5 second 720p clip usually takes about 3.5 minutes.

Pass up to 10 images in reference_images, as https URLs or base64 data URLs of PNG, JPEG or WebP files up to 10 MB, and mention them in the prompt as @Image1, @Image2 and so on, in the order you sent them. Say what each is for, such as "keep the face and jacket from @Image1" or "the kitchen from @Image2"; without that, the model guesses. Reference images can't be combined with a first or last frame.

Yes. Pass first_frame to animate an image, and add last_frame to move between two images. The video takes the first frame's shape. Images can be https URLs or base64 data URLs of PNG, JPEG or WebP files up to 10 MB.

Always. Every clip comes with generated stereo sound, such as dialogue, ambient sound, music and effects, and the price includes it. Write lines in quotes next to the character who says them, and name the sounds you want. For a silent video, mute or strip the audio track, or use Kling 3.0, whose silent clips cost less.

720p and 1080p, in 16:9, 9:16, 1:1 or 21:9. From a first frame, the video takes the image's shape; with reference images, set aspect_ratio to adaptive to let the prompt or the images decide. An OpenAI-style size such as 1920x1080 works too.

Not yet. On Zurelay, Kling 4.0 takes a prompt, a first and last frame, or up to 10 reference images. For reference videos, editing and extending, use Seedance 2.5 or Gemini Omni Flash.

Kling 4.0 makes clips up to 30 seconds instead of 15, takes up to 10 reference images, adds 21:9, and always makes sound. Kling 3.0 costs less per second, can make cheaper silent clips, and takes a negative prompt.

Yes. Your requests run on Kuaishou's Kling 4.0, and Zurelay doesn't swap in a different or smaller model. Video generation varies from run to run, so the same prompt makes a different clip each time. Zurelay is an independent service and is not affiliated with Kuaishou.