Kling V4 AI Video Generator
Kling V4 makes 3 to 30 second clips at 720p or 1080p from a prompt, a first frame, or a set of reference images, videos, and audio.
Keep building
Key Features
Three ways in: write a prompt, animate a first frame, or point the model at reference images, videos, and audio.
Clip length from 3 to 30 seconds, in one-second steps.
720p or 1080p, in 16:9, 9:16, 1:1, or 21:9.
Start and end frames, plus middle keyframes, so a clip can pass through the pictures you choose.
Reference mode takes up to 10 images, 5 videos, and 3 audio clips. Name them in the prompt as @Image1, @Video1, and @Audio1.
Stereo audio is part of every clip. There is no separate audio switch.
How to use Kling V4
- Step 1
Pick the input that matches the job
A prompt alone is enough for a new shot. Add one image when that picture should be the first frame. Add more images, a video, or audio when you want the model to follow a subject, a motion, or a voice.
- Step 2
Say what moves
Lead with the subject, the action, and the camera. 'A ranger crosses a torchlit bridge, slow dolly in, rain on the stone' is more useful than a broad scene name.
- Step 3
Use 720p while you look
720p is the cheaper rate. Move to 1080p once the motion and the framing are right.
- Step 4
Point at references by number
In reference mode the prompt can name @Image1, @Video1, and @Audio1 in the order you uploaded them.
Best use cases
Scene openers
Short establishing shots and character entrances for a session or a recap.
Directed motion
A start frame and an end frame when the clip has to arrive at a picture you already have.
Kept subjects
Reference images and a short video when the same person, creature, or camera move has to survive the shot.