Portrait Image
PNG, JPG, WEBP
Click or drag in an image. A clearer, less obstructed face improves generation stability.
AI portrait animation
LivePortrait Online Portrait Animation Generator
Upload a clear portrait image and a driving video with facial expression or head movement to generate a natural animated portrait online.
- Each visitor works in an independent browser session
- Results include original-space and cropped animation videos
- Upload your own media or test quickly with official examples
Step 1
Upload Media
Use a front-facing or three-quarter portrait. A near-square driving video focused on the head usually works best.
Driving Video
MP4, WEBM, MOV
The first frame should show a natural front-facing pose. Keep shoulder movement modest and prefer short clips.
Optional
Example Media
Select a portrait image or driving video independently, or pick one of each official example to test the flow.
Source Portrait Examples
Click any image to replace only the portrait image.
Driving Video Examples
Click any video to replace only the driving video.
Step 2
Animation Options
The default options work for most portraits. Keep paste-back enabled when you want to preserve the original composition.
Processing Status
Waiting for UploadUpload a portrait image and a driving video to start generation.
Remote queue and processing progress will appear here.
Step 3
Generated Results
After generation finishes, preview or download the video files directly.
The original-space video will appear here
The cropped animation video will appear here
Feature Overview
Turn a static portrait into a natural animated video
This lightweight client keeps the core LivePortrait flow while offering clearer upload, progress, and result areas.
Photo-driven Animation
Use a portrait image as the subject and a video to provide expression, mouth, and head movement.
Two Result Previews
View both the original-space video and the cropped video so you can download the version that fits your use case.
Independent Browser Sessions
Each visitor's file selections, progress, and results stay in the current browser page state and do not reuse another user's session.
Responsive Page
The upload, example, option, and result areas adapt to desktop and mobile screen widths.
User Guide
Suggested workflow for stable results
- Upload a clear portrait image. A visible face, even lighting, and a subject that is not too small work best.
- Upload a driving video. A 1:1 frame with the head centered and a natural front-facing first frame is recommended.
- Keep the default options and click "Start Generation". If the face is small in the image, keep "Auto-crop Portrait" enabled.
- Wait for the remote GPU queue to finish. Busy queues can take longer, and the page will keep showing the current status.
- When generation completes, preview both videos in the result area and download the version that best fits your needs.
Media Tips
To reduce failures, use short driving videos and avoid large turns, fast motion, heavy occlusion, or multiple people in the frame. Confirm you have the right to use any public videos or portrait media.
FAQ
Know these limits before generating
When are uploaded files sent?
The selected portrait image and driving video are sent to the remote LivePortrait service only after you click "Start Generation".
Why can the queue or wait time be long?
This feature depends on a remote GPU. Queue and processing times increase when traffic is high, the Space is waking up, or the media files are large.
Does each user have an independent client?
Yes. The page runs independently in each visitor's browser, does not use shared local state, and does not show one user's selections to another user.