How long can a Kling 4.0 video be?
Choose 3–30 seconds in whole seconds. Edit video can follow the source duration. Extend video adds a 3–30-second segment and returns it joined to your original video.
Text · Key frames · References · Edit · Extend
Create 3–30-second videos from text, key frames and reference media. Edit or extend an existing video, with 720p / 1080p output and stereo sound.
Workspace
Kling 3.0 / Turbo reference briefs
A single five-second cinematic medium shot at shoulder height. A woman in a mustard-yellow raincoat pauses beside a warmly lit window on a rainy evening, then gently turns her face toward the light. The composition includes only her upper body, softly blurred amber light through plain glass, and rain in the foreground. No storefront facade, signs, lettering, readable books or street names are in frame. Slow restrained camera push-in, natural blinking, consistent facial features, realistic raindrops and warm reflections, no cuts. Audio: soft falling rain and a quiet distant city ambience, no speech or music. No text, logos or subtitles.
Use the uploaded perfume photograph as the first frame. The camera slowly slides a short distance to the right, keeping the single amber glass bottle and rounded black cap in focus. A soft daylight reflection travels gently across the glass. Keep the bottle upright on the walnut plinth, with its cap closed and its silhouette unchanged. Preserve the warm sand-colored wall and natural contact shadow. One continuous shot, no cuts, no added objects, labels or text.
A lone adult cyclist wearing a rust-orange jacket rides a plain dark bicycle from left to right along a flat coastal path at sunrise. A low side-tracking camera moves alongside at the same pace, keeping the entire bicycle and rider in view. Distant sea and pale grasses move gently behind them. Natural soft light, realistic proportions, restrained documentary style. One continuous shot with steady forward motion, no cuts, no logos or on-screen text.
A handcrafted paper observatory stands on a small indigo hill beneath a deep blue evening sky. Warm amber light glows inside its round windows. Layered paper pine trees frame the foreground, with small folded clouds farther away. The camera slowly pushes forward between the trees toward the observatory, revealing gentle parallax between the paper layers. Keep the observatory shape and window positions stable. Tactile cut-paper edges, soft miniature-set lighting, one continuous shot, no people, lettering or cuts.
Starting from the provided room image, a light breeze slowly lifts the sheer curtain beside the open window and lets it settle. The camera stays locked. Keep the window frame, furniture, wall details, and daylight direction unchanged. One continuous shot, no new objects or text.
One continuous low-angle tracking shot of a single blue paper sailboat floating from left to right across a shallow sunlit pond. The camera gently follows at water level as the boat passes a reed. Small ripples trail behind it. Preserve the folded paper hull and sail, soft afternoon reflections and realistic scale. No people, text, cuts or additional boats.
Kling 4.0 turns a written scene, key frames or reference media into a video with stereo sound. Start with text for a new shot, use images to guide its composition, or bring existing footage into an edit or continuation.
Choose 720p or 1080p, and describe the action and sound together. A short first draft helps you check the subject, camera move and timing before committing to a longer scene. For another Kling workflow, open Kling 3.0 or Kling 3.0 Turbo.
| Mode | What to prepare | Use it for |
|---|---|---|
| Text to video | A prompt; 3–30 seconds | A new scene from a written brief |
| Key frames | One first frame; optional end and up to 8 middle frames | Guiding the visual progression of a shot |
| References | Images, video or audio; up to 15 files in total | Combining appearance, movement and sound references |
| Edit video | One source video of 3–30 seconds; optional reference images | Changing a scene while using existing footage |
| Extend video | One source video up to 30 seconds; optional end frame | Continuing the action with a new 3–30-second segment |
Upload JPG, PNG or WebP images under 30 MB, MP4 or MOV videos under 50 MB, and MP3 or WAV audio under 15 MB. Reference videos share a 30-second total; audio clips are individually 5–30 seconds. Only upload media you have permission to use.
Use Kling 4.0 when your scene needs more than 15 seconds, middle key frames, or the edit and extend workflows on this page. Kling 3.0 and Turbo remain separate workspaces with their existing controls and pricing.
| Here on RSW | Kling 4.0 | Kling 3.0 | Kling 3.0 Turbo |
|---|---|---|---|
| Clip duration | 3–30 seconds | 3–15 seconds | 3–15 seconds |
| Frame controls | First, end and up to 8 middle frames | First and end frames | First frame |
| Aspect ratios | 16:9, 9:16, 1:1, 21:9; follow input where supported | 16:9, 9:16, 1:1 | 16:9, 9:16, 1:1 |
Use Text for an original scene, Key frames for an image-led transition, or References to combine media. Edit and Extend each start from one source video.
Name the subject, one main action, the camera move and the sound. In References, type @ to insert Image, Video or Audio references in the order shown.
Review duration, resolution and credits. Follow the task in History or My Creation, then check the result before downloading or recreating it.
For a product shot, describe how the camera reveals the object and what stays still. For dialogue, identify the speaker and put the line in quotes. Review hands, labels, voice timing and continuity in the finished clip. Reference media guides a generation; it does not guarantee an exact reproduction.
Create with Kling 4.0Choose 3–30 seconds in whole seconds. Edit video can follow the source duration. Extend video adds a 3–30-second segment and returns it joined to your original video.
This workspace offers 720p and 1080p. Generated videos always include stereo audio; there is no audio-off option. Edit video lets you preserve source audio or generate new audio.
References mode accepts up to 10 images, 5 videos and 3 audio clips, with 15 files in total. Reference videos must total no more than 30 seconds. Each audio clip must be 5–30 seconds.
Add one first frame, an optional end frame and up to 8 middle frames. The combined limit is 10 images. Middle frames follow the order shown in the editor. A prompt is optional in this mode.
Kling 4.0 uses credits and is excluded from Unlimited. The Generate button shows your quote before submission. Video edits and generations with a reference video account for both input and output duration. Extensions are charged for the new segment.
The reference briefs shown in Inspirations are existing Kling 3.0 and Kling 3.0 Turbo clips, labeled with their source model. They demonstrate prompts you can adapt, rather than Kling 4.0 results. Using a prompt only fills the editor; it does not generate or spend credits.