Generate Up to 30 Seconds in One Take
Kling 4.0 generates up to 30 seconds in a single pass, twice the 15-second limit of Kling 3.0. Fit a beginning, middle and end into one continuous shot — no stitching, no jump cuts.
Start a 30s VideoKling 4.0 creates 30-second videos with stereo audio, supports 15 multimodal references and gives you 10-keyframe control to shape your story. Try Kling 4.0 on ChatArt now.
Real results from Kling 4.0. Hover to pause, click to start from the same prompt.
Product film · 15.0s · 16:9
Beauty editorialUse prompt
Product ad · 30.0s · 16:9
Tropical soda adventureUse prompt
Animation · 14.4s · 16:9
Masked heroine vs. monstersUse prompt
Action · 13.8s · 16:9
Ice and fire showdownUse prompt
Animation · 15.0s · 4:3
Dragon rider above the cloudsUse prompt
Showreel · 13.1s · 16:9
UFO chase and cinematic scenesUse prompt
Kling 4.0 is the latest AI video model from Kling AI (Kuaishou), officially launched on October 15, 2026. It generates up to 30 seconds of video in a single pass from images, video clips and text. Compared with Kling 3.0, it doubles the maximum length, adds up to 10 keyframes and 15 reference assets, keeps fast motion more stable and outputs two-channel stereo audio.
On ChatArt, Kling 4.0 sits next to Seedance 2.5, Veo 3, Wan 3.0 and other leading models — one account, one workspace. Every video can be used commercially and downloads with no watermark.
Start a 30-second video →Five upgrades that matter when you are making real videos, not just short clips.
Kling 4.0 generates up to 30 seconds in a single pass, twice the 15-second limit of Kling 3.0. Fit a beginning, middle and end into one continuous shot — no stitching, no jump cuts.
Start a 30s Video
00:00
00:08
00:12
00:30
Pin images to the moments that matter. Kling 4.0 fills in the motion between them, so second 12 looks exactly the way you planned.
Set your keyframes →
Add up to 10 images and 5 video clips, plus voice, so the same character, outfit or product stays recognizable from first frame to last.
Add your references →
Chases, sports and big camera moves stay stable, with smoother continuity during fast motion and complex camera work.
Try an action scene →
Footsteps, engines, rain and dialogue are generated in two-channel stereo with the picture, with lip sync that holds up across languages and accents.
Make a talking video →Keep the bottle, box or logo consistent with product references and land on a clean hero shot.
Tell a full beginning, middle and end in one 30-second take, with the same characters throughout.
Make vertical 9:16 clips for TikTok, Reels and Shorts that hold attention past the first second.
Create characters who speak with stereo sound and natural lip sync, in English, Japanese, Korean and more.
Walk viewers through a product or idea, using keyframes so each step lands exactly on time.
Cut action to the beat, then add a soundtrack with AI Music Video.
Watch the result, then start from the same prompt in one click.
15.1s · 9:16Vertical 9:16, 24fps, 15 seconds. Vibrant golden ingredient Money Shot, flower-to-honey visual journey, no people, macro natural textures, high-energy but controlled studio transitions. Use image 1 as the exact bear-shaped honey bottle. Preserve molded bear face and body, translucent plastic, golden fill, label, narrow neck, and dark-blue ribbed cap.0-2s | Nectar macro: a single golden nectar droplet rests on a flower stamen in warm sunlight; camera circles slowly.2-4s | Golden release: the droplet expands into a controlled cloud of amber liquid ribbons and tiny pollen-like particles, never explosive or messy.4-6s | Bottle emergence: the exact filled bear bottle rises from the amber cloud, molded face and label forward, condensation-free plastic catching rim light.6-8s | Source detail: macro cuts of honeycomb cells, one clean honey strand, and the bottle's molded paws and ribbed cap.8-10s | Real opening: hands-only shot unscrews the dark-blue cap from the threaded neck; no flip-top is invented.10-12s | Pour proof: bottle tilts and one slow golden stream pours from the open neck onto a spoon, then cleanly stops.12-15s | Hero: recapped bear bottle centered with restrained honeycomb geometry behind it and one amber light halo.Audio: soft liquid stretch, cap twist, pour, warm ambient rise. Negative Prompt: flip cap, squeeze nozzle, changed bear face, fruit bottle, multiple flavors, bees swarming, sticky mess, faces, subtitles, watermark.
8.1s · 3:4A continuous one-shot sequence: rapidly zoom the camera out at extremely high speed with dynamic motion blur, ensuring smooth transitions that follow natural zoom progression. First, reveal a top-down view covering an area of 100 km. Once the previous frame shrinks out of sight, continue zooming out through thick layers of clouds to reveal a 1000 km view. After the prior frame disappears, show the entire country or continent. Finally, when the last frame becomes invisible, reveal the entire Earth and end the sequence. The Earth should remain stationary (no rotation).
8.1s · 3:4The subject remains completely still as its entire body disintegrates into a massive cloud of gray ash, filling every corner of its form. Slowly, the wind rises, carrying both the subject and its fragments away. One by one, they vanish into the void, leaving behind only an eerie silence and an overwhelming sense of absence. The space once occupied by the body is now empty, echoing the silence of its loss.
Based on Kling AI's official release notes. Both models are on ChatArt.
| Feature | Kling 3.0 | Kling 4.0 |
|---|---|---|
| Max video length | 15 seconds | 30 seconds, single pass |
| Keyframes | Start and end frame only | Up to 10 keyframes |
| Reference assets | Up to 7 | Up to 15 (images, video, voice) |
| Dynamic range | 8-bit SDR | 10-bit HDR at 1080p and 4K |
| 21:9 ultrawide | Not available | Supported |
| Audio | Single-channel | Two-channel stereo |
You want a full 30-second scene, need to pin up to 10 keyframes, or want stereo dialogue with lip sync in a 21:9 frame.
You only need a quick clip under 15 seconds. See Kling 3.0
You want to guide a video with lots of references — up to 50 — or edit specific regions of a shot. See Seedance 2.5
Kling 4.0 is a big step up, but a few details still deserve a second look.
Detailed hand work can still drift. Frame hands a little wider, or regenerate that shot.
Put exact words in quotes and keep them to a few words for the cleanest result.
For a story told across several videos, use the same voice reference so the character sounds the same.
Upload two to four angles of a product so labels and shapes stay accurate.
Open Image to Video on ChatArt and select Kling 4.0 from the model list.
Upload up to 15 references or 10 keyframes, describe the scene, then choose a length up to 30 seconds and an aspect ratio.
Download with no watermark, or keep going with Video Extension to make it longer.
More questions? Contact support
Kling 4.0 is Kling AI's newest video generation model, launched on October 15, 2026. It creates up to 30 seconds of video in a single pass from images, video clips and text, with up to 10 keyframes, up to 15 reference assets and two-channel stereo audio. You can use it on ChatArt in your browser.
Kling 4.0 doubles the maximum length to 30 seconds, supports up to 10 keyframes instead of only a start and end frame, accepts up to 15 references instead of 7, adds 10-bit HDR and 21:9 output, and upgrades audio from single-channel to two-channel stereo.
Kling 4.0 generates videos from 3 to 30 seconds in a single pass, twice the 15-second limit of Kling 3.0. On ChatArt you can generate the full 30 seconds, and use Video Extension when you need an even longer clip.
Yes. Kling 4.0 generates two-channel stereo audio in the same pass as the video — sound effects, ambience and dialogue — with more accurate lip sync across languages and accents. Describe the sounds you want at the end of your prompt.
Kling 4.0 supports 16:9, 9:16, 1:1 and 21:9. Use 9:16 for TikTok, Reels and Shorts, 16:9 for YouTube and websites, 1:1 for social feeds and 21:9 for a cinematic widescreen look.
Yes. On ChatArt, open Image to Video and select Kling 4.0. Use frames-to-video mode to animate between up to 10 keyframes, or reference mode to build a scene from up to 15 files — up to 10 images and 5 videos.
Write it in beats. Describe the subject, then what happens at 0–10s, 10–20s and 20–30s, add the camera move and lighting, and end with the sounds you want. Put any exact dialogue or on-screen text in quotes.
Yes. Videos you make with Kling 4.0 on ChatArt can be used in commercial projects such as ads, product videos and client work, and they download with no watermark. See our Terms & Conditions for details.
Upload your references, describe the scene, choose 30 seconds and hit Generate.