Motion Control AI beginner guide

Make your first motion-controlled video without guessing

You only need two files: a picture of the character and a video of the movement. This guide explains every setting in plain English.

Open the generator

Before you start

Prepare one file for appearance and one file for movement. The AI combines them.

Character image = who appears

Use a bright, clear JPG or PNG. Show the head, shoulders, torso, and any body parts needed for the movement. Avoid crossed arms, cropped limbs, and objects covering the body.

Movement video = what the character does

Use a 3–30 second MP4 or MOV with one continuous action. Keep the moving person fully visible, well lit, and free from obstructions.

Four steps from files to finished video

Step 01

Choose a model

Start with Kling 2.6 Pro to use fewer credits. Choose Kling 3.0 Standard when you want higher-fidelity motion transfer.

Step 02

Upload your character

Choose a clear JPG or PNG showing the person, illustration, avatar, or mascot you want to animate. Keep the body visible and unobstructed.

Step 03

Upload the movement

Choose an MP4 or MOV showing the dance, gesture, walk, or action to copy. A short clip with one clearly visible person gives the most reliable result.

Step 04

Check and generate

Pick which framing to follow, decide whether to keep sound, add an optional style prompt, then review the credit cost and generate.

Which model should you choose?

Recommended starting point

Kling 2.6 Pro

Choose this for first attempts, prompt testing, and lower credit use. It costs 1.6 credits per detected video second.

For shots that need extra fidelity

Kling 3.0 Standard

Choose this for higher-fidelity motion transfer. It costs 1.8 credits per detected video second.

Framing, without the jargon

This setting tells the AI which camera angle and composition should guide the result. It does not change where the character comes from.

Match the movement video

Use this default when you want the character to face and move like the person in the motion clip. It supports clips up to 30 seconds.

Keep the character image

Use this when the original image composition matters most. With Kling 3.0 Standard, the movement video must be 10 seconds or shorter.

Upload checklist

Character: JPG or PNG
Movement: MP4 or MOV
Maximum 10 MB per file
At least 300 px on both sides
Aspect ratio between 1:2.5 and 2.5:1
Movement clip between 3 and 30 seconds

What should the prompt say?

Describe the scene, lighting, atmosphere, and style. Do not repeat the dance or action—the video already provides it.

Good prompt

Cinematic lighting, golden hour, urban street background, shallow depth of field, subtle film grain.

Negative prompt

Blurry, distorted body, extra limbs, watermark, low quality, flickering.

Leave “Keep Original Sound” on to copy the movement video's audio. Turn it off for silent output.

Common problems and the next thing to try

My file is rejected

Use JPG/PNG for the character and MP4/MOV for the motion. Keep each file under 10 MB and at least 300 px on both sides.

The movement looks weak

Use a clip where the moving person stays fully visible. Avoid cuts, crowds, objects covering the body, and very dark footage.

The body becomes distorted

Choose a character image that shows a similar body area and angle to the person in the motion video. Shorten complex dance clips to 3–10 seconds.

The framing is not what I expected

Choose “Match the movement video” for the safest result. Use “Keep the character image” only when preserving the image composition matters more.

The result has no sound

Turn on “Keep Original Sound” before generating. This copies the audio from your uploaded movement video.

Generation is taking a long time

Keep the page open. Motion Control can take several minutes, and longer clips usually need more processing time.

Beginner questions

No. The uploaded video already tells the model which movement to copy. Use the prompt for the background, lighting, atmosphere, and visual style.

Start with Kling 2.6 Pro because it uses fewer credits. Move to Kling 3.0 Standard when a shot needs higher-fidelity motion transfer.

Use a clip from 3 to 30 seconds. Kling 3.0 Standard is limited to 10 seconds when you choose the character-image framing option.

Credits follow the detected video length. Kling 2.6 Pro uses 1.6 credits per second and Kling 3.0 Standard uses 1.8 credits per second, rounded up.

Ready to make the character move?

Start with one clear image and one short movement clip. You can improve the style after the first result.

Create a Motion Control Video