Longer shots, finer control
Kling 4.0
Kuaishou’s next video model is built for longer shots and tighter story control.
Create a video
Write a prompt or add a start image, then choose duration, aspect ratio, and resolution.
0 / 2000
What Kling 4.0 adds
Kling 4.0 is built for longer clips and more precise control over the story.
Native 30-second clips
A single generation runs about 30 seconds, longer than the 3–15 second range on Kling 3.0.
Up to 15 references
Coverage describes a mix of images, video clips, and subject references, reported as 10 images, 5 videos, and 7 subjects.
Up to 10 keyframes
Story beats can be pinned across the clip, beyond a single start frame and end frame.
4K and 10-bit HDR
Up to 4K and 10-bit HDR, with clearer detail in highlights and shadows.
Stereo sound and lip sync
Native audio, clearer lip sync, and speech in several languages, accents, and dialects.
Longer extensions
Repeated extension is reported toward about two minutes. A 21:9 frame is also mentioned in coverage.
Closest models you can use now
Seedance 2.5 is the closest match for a 30-second clip with image and video references. Kling 3.0 is the previous generation in the same line.
| Kling 4.0 | Kling 3.0 | Seedance 2.5 | |
|---|---|---|---|
| Length | About 30 seconds | 3–15 seconds | Up to 30 seconds |
| References | Up to 15 | First and last frame, plus element references | Images, video, and audio references |
| Keyframes | Up to 10 | Start and end | Start, end, and references |
| Audio | Stereo and lip sync | Native audio | Native audio |
How to generate on this page
- Choose a model in the form.
- Write a prompt, or add a start image for image-to-video.
- Choose duration, aspect ratio, and resolution, then generate.
Kling 4.0 questions
How long is a Kling 4.0 clip?
A native clip is about 30 seconds, and it can be extended further.
How many references can Kling 4.0 use?
Up to 15, mixing images, video clips, and subject references.
How does keyframe control work?
Up to 10 keyframes can pin story beats across the clip, beyond a single start frame and end frame.
Which model is closest to Kling 4.0?
Seedance 2.5 is the closest match for a longer clip with image and video references and native audio. Kling 3.0 is the previous generation in the same line.
How does Kling 4.0 differ from Kling 3.0?
Clips run longer, about 30 seconds instead of 3–15. It also takes more references and keyframes, with stereo sound and clearer lip sync.