How It Works
1. Upload your Image
Choose a clear portrait or full-body photo of your character.
Supported formats: JPG or PNG.
The image should show a well-lit, unobstructed face (no sunglasses or strong shadows).
Recommended minimum: 512px per side.
2. Upload your Video
Add the motion source — the video that contains the pose, gestures, and expressions to be transferred.
Duration and frame rate will be taken directly from the video.
The minimum charged duration is 5 seconds.
Supported formats: MP4, MOV.
3. Choose a Mode
Replace — swaps the person in the video with your person from the uploaded image.
Animate — animates your photo using the same movements and expressions as in the video.
4. Set Resolution
Select 480p for faster results or 720p for higher quality output.
5. Generate
Click Generate and wait a moment while the model processes your clip.
Your result will appear in the task list — ready to preview, download, or send downstream to Publishing.
Tips for Best Results
Match Composition & Pose
Keep the uploaded image and video aligned in camera position and body pose.
Misalignment can lead to facial distortion or frame mismatches.
Keep Aspect Ratios Consistent
If your photo and video have different aspect ratios, the AI must crop or stretch to fit.
Matching ratios gives more accurate face mapping and smoother motion.
Quality Matters
The higher the resolution and clarity of your inputs, the better the output.
Avoid blurry faces, overexposed lighting, or extreme motion blur.
Output
Duration: taken from your uploaded video
Resolution: 480p or 720p
Format: MP4
Processing time: depends on video length and GPU queue
Common Questions
What happens if my video doesn’t contain a visible face?
You’ll receive an error — but your credits will be automatically refunded.
Can I upload long clips?
Yes, up 120 sec.