Sign in to see your generations here
Kling 4.0 by Kuaishou is a native multimodal video model built for directors and storytellers who need more than a single continuous shot. It accepts text, still images, reference clips, and even a designated end frame, then renders up to 30 seconds at 720p or 1080p with per-shot prompting that keeps characters and environments consistent across cuts. It suits brand teams, animators, and content creators who want structured, cinematic output rather than a single unbroken take.
How to use this app
- 1
Describe what you want to create, or upload your source file.
- 2
Pick your options, then press Generate.
- 3
Watch your result appear in the gallery within moments.
- 4
Download, share, or generate again with a new idea.
What you can make
Multi-Shot Brand Films
A product launch needs an opening close-up, a lifestyle mid-shot, and a logo reveal. Kling 4.0's per-shot prompting lets you script each beat separately while the model holds the product's colour and form consistent from one shot to the next.
Character-Driven Short Stories
Supply a character reference image and write scene-by-scene prompts for a brief narrative. Kling 4.0's reference consistency engine keeps facial features and costume details stable across shots, avoiding the drift that breaks immersion in longer storytelling.
Controlled Start-to-End Sequences
Feed Kling 4.0 a starting frame and a separate end frame to define the boundaries of a transition or reveal. The model interpolates motion between them, giving you predictable arc-based animation without manual keyframing.
Complex Motion Scenes
Action sequences, crowd movement, and camera crane moves benefit from Kling 4.0's improved stability through complex motion. Fast pans and layered movement that cause jitter in lighter models render cleanly at up to 1080p.
Reference-Driven Location Work
Attach reference clips of a real or stylised environment and Kling 4.0 will match its visual language across shots, useful for travel content, architectural walkthroughs, or world-building in a consistent fictional setting.
Prompt ideas to try
Why creators use this app
- Per-shot prompts
- Reference to video
- Start and end frame
- Up to 30s
Tips for better results
Use Per-Shot Prompts for Structure
Rather than describing an entire sequence in one prompt, break your vision into distinct shot descriptions and let per-shot prompting handle each. This gives the model clearer direction and produces more intentional cuts rather than one long wander.
Anchor Characters with Reference Images
Kling 4.0 is built for reference consistency, so provide a clean, well-lit character image as your reference input before writing scene prompts. The model will carry that likeness forward across shots far more reliably than text description alone.
Set Duration Before Prompting
Choose your clip length using the Duration control before you write your prompt. A 5-second clip rewards tight, single-moment descriptions; a 10-second clip can sustain a two-beat scene with a clear transition built into the language.
Match Aspect Ratio to Your Platform
Use the Aspect Ratio control to pick the correct frame shape for your destination. Select 9:16 for vertical social content, 16:9 for standard video, and 21:9 for a cinematic widescreen look before generating, so the composition is designed for that frame from the start.
When to choose this app
Choose Kling 4.0 when your project requires structured multi-shot storytelling, stable character identity across cuts, or start-and-end-frame control over motion arcs. Seedance 2.5 and Seedance 2.0 both bring native audio synthesis to their output, which Kling 4.0 does not prioritise. Wan 3.0 offers single-pass omni-reference video generation. Kling 4.0 is the better fit when per-shot scripting and reference consistency across a cast or location matter more than built-in audio.
Frequently asked questions
What does per-shot prompting mean in practice?
Instead of writing one continuous description, you write a separate prompt for each planned shot in your sequence. Kling 4.0 uses those individual instructions to shape the motion, framing, and mood of each segment while still maintaining visual consistency between them, much like a shot list for a real production.
Can I supply more than one reference image?
Yes. Kling 4.0 accepts a set of reference images and clips as input, not just a single still. You can provide images of a character, a location, and a stylistic reference simultaneously, and the model will draw on all of them when composing the video.
How does the start and end frame feature work?
You supply a still image as the opening frame and a separate still as the closing frame. Kling 4.0 generates the motion between those two visual states, giving you a controlled transition or reveal where both the beginning and the end point of the shot are exactly what you specified.
What resolution will my video be?
Kling 4.0 renders at either 720p or 1080p. The output resolution depends on the settings applied at generation. For finished deliverables intended for large screens or professional editing timelines, 1080p provides the most headroom for cropping and colour work.
How long can a single generated clip be?
Kling 4.0 supports clips from 3 to 30 seconds. The Duration control on the page currently offers 5-second and 10-second presets. For longer sequences, plan to chain multiple clips using consistent reference inputs so character and scene details carry across generations.
Is Kling 4.0 suitable for animating non-photographic styles?
Yes. Because the model accepts reference images as style anchors, you can supply illustrated, painted, or 3D-rendered stills and it will attempt to match that visual language. Results are strongest when the reference is clear and internally consistent rather than a collage of unrelated styles.
Which AI model powers this app?
This app runs on Kling 4.0, available through Arteza with no separate account or setup.
Can I use the results commercially?
Yes. Content you generate is yours to use, subject to our content licenses.
How long does a generation take?
Most generations finish in under a minute, and you can watch progress live in the gallery.