Soracai LogoSoracai.com
Back to Home

AI Motion Control

Upload your own motion video and a photo to create custom AI videos. Your character, your moves!

1. Upload Motion Video

Upload a video with the motion you want to apply (max 30s)

2. Choose Your Character

AI will first generate a new character in the video's scene, then apply the motion

3. Generate Video

Motion Video:Not selected
Character:Not selected
Keep Background:Yes

How it works

  • 1.Upload a motion reference video (3-30 seconds, MP4/MOV/MKV)
  • 2.Upload a photo of your character or describe them with a prompt
  • 3.Toggle "Keep Background" to preserve the original scene
  • 4.AI generates a video with your character performing the motions

Create Custom Motion Videos with AI

AI Motion Control lets you create videos where any person performs any motion you choose. Unlike preset dance templates, you upload your own reference video — whether it's a dance, workout routine, sign language, or any other movement — and AI applies that exact motion to your chosen character.

Choose between uploading a photo of the character or describing them with a text prompt. With the "Keep Background" option, you can even preserve the original scene from your motion video while swapping in a different character.

Two Ways to Define Your Character

Upload a Photo: Provide a clear photo of the person you want in the video. The AI will transfer the motion from your reference video onto the person in your photo, creating a seamless result.

Use a Text Prompt: Describe the character you want — their appearance, clothing, and style. The AI will generate a character matching your description and apply the motion to them.

Keep the Background Feature

When "Keep Background" is enabled, the AI first analyzes the first frame of your motion video and creates a new version with your character placed in the same scene. This means the background, lighting, and environment stay the same — only the person changes. The result is then used as the starting point for motion control, creating a natural-looking video where your character performs the motion in the original setting.

How to Create Motion Control Videos

1Upload Your Motion Video

Record or select a video with the motion you want to transfer. This can be a dance, exercise, gesture, or any movement. Duration must be between 3 and 30 seconds. Supported formats are MP4, MOV, and MKV (max 100MB). The video should clearly show the subject's head, shoulders, and torso for best results.

2Choose Your Character

Upload a JPEG or PNG photo (min 300px, max 10MB) of the person you want in the video. The photo should clearly show head, shoulders, and torso. Alternatively, use a text prompt to describe the character. For prompts, be descriptive about appearance and clothing.

3Configure Options

Toggle "Keep Background" to preserve the original scene from your motion video. When enabled, the AI will place your character into the video's environment before applying the motion.

4Generate & Download

Click generate and wait for your video. Generation takes 2-10 minutes depending on whether the keep-background option is used. Download your video when ready and share it on social media.

Frequently Asked Questions

What's the difference between AI Dance and Motion Control?

AI Dance uses preset dance templates — you pick from 23+ styles. Motion Control lets you upload your own motion reference video, giving you complete creative freedom. Any movement you can record becomes available as a motion template.

What does "Keep Background" do?

When enabled, the AI preserves the background and environment from your motion video. It first swaps the character in the scene using AI image generation, then applies the motion. This creates a more natural result where your character appears in the original setting.

How long can my motion video be?

Motion reference videos must be between 3 and 30 seconds long. Supported formats are MP4, MOV, and MKV (max 100MB). Shorter videos (5-15 seconds) tend to produce the best results. The video should clearly show the subject's head, shoulders, and torso with clear, visible movements.

Can I use a text prompt instead of a photo?

Yes! You can describe the character you want using a text prompt. The AI will generate a character matching your description and apply the motion to them. Be specific about appearance, clothing, and style for the best results.

How long does generation take?

Direct mode (photo without keep-background) takes 2-5 minutes. When using keep-background or a text prompt, it takes 5-10 minutes because the AI first generates the character photo, then creates the motion video.

How much does it cost?

Each motion control video costs 10 coins regardless of the mode used. You can purchase coin packages starting from $5.99 for 10 coins.