Mixer AI
Kling 3.0 Motion Control
Transfer a recorded performance to your character and choose which source supplies the background.
When to choose it
Kling 3.0 Motion Control replaces the performer rather than inventing the performance. Video supplies movement; an image supplies appearance. You can choose the photo or video's background separately. Use it for dances, gestures and scenes where matching an existing performance matters most.
- Show the desired performance through a reference clip.
- Choosing the background source helps place the character in the intended setting.
Modes and settings
- One character image and one motion video; 720p or 1080p.
- Video orientation supports up to 30 seconds; image orientation up to 10. The background can come from either source.
Price in coins
| Quality / size | Input | Coins / second |
|---|---|---|
| 720p | Photo + motion video | ≈ 13.5 |
| 1080p | Photo + motion video | ≈ 17.91 |
Billing is per second, not per fixed-length clip. The rate is multiplied by the billable duration; the complete request total is rounded up to a whole coin once.
Billing uses the duration of the motion video transferred to the character.
≈ marks a fractional rate rounded to two decimal places for display; billing uses full precision.
What you can make
- Place an illustrated character into a filmed dance scene.
- Repeat a presenter's gestures using a different prepared character.
How to direct the result
- Match framing and gaze direction between the photo and the video's opening.
- Get motion working first. Avoid changing costume, background, lighting and body shape together.
Example: repeat a presenter's gestures
The character in the photo repeats the movements in the video. Preserve the character's costume and proportions, using the video's background. Calm gestures with both hands fully visible. Do not add props to the hands or other people to the scene.
What to keep in mind
- A motion video is required; a dance description alone is not enough.
- Clothes and faces can change during occlusions. Inspect intermediate frames, not just the opening.
Sources
- Prices and availability — Mixer AI catalog, updated 2026-09-18.
- Official maker site: Kuaishou
Model facts verified: 2026-09-20.
Guides and comparisons using this model
FAQ
How is this different from regular Kling generation?
Movement comes from an uploaded performance instead of being invented from text alone.
Can I keep the photo's background?
Yes. Choose the photo as the background source, or use the video background.
Should the inputs have similar poses?
Similar angles and framing help, particularly for full-body movement and wide gestures.
Why are hands cropped?
The photo may lack space, or the video may crop the performer. Prepare a wider frame.
Can I combine several motion videos?
One request uses one clip. Assemble the required performance into a single video first.
Should I test at 1080p?
Check pose and motion at 720p first. Higher resolution does not fix incompatible source material.
What changes the price?
The coin rate per second depends on quality, mode and, for some models, sound. Multiply it by the billable duration. Where input video is charged, its duration also counts. The complete request total is rounded once.
Do I need a separate model subscription?
No. In Mixer AI, you top up one coin balance and use it for the tasks you need. You do not need a separate subscription to this model.