Free VideoReTalking Generator
Use VideoReTalking with an existing face video and a new speech track to generate synchronized mouth movements while retaining the speaker and scene.
Try the tool
What you need
- Video
- Speech audio
Re-dub a Talking-Head Video
Loading the demo…
How to use it
1. Prepare your inputs
Open the demo and add the inputs listed above. You can start with an example inside the demo before uploading your own files.
2. Set up the result
Choose one clearly visible face, clean speech and a short clip. Match the audio to the segment you want to re-dub, then inspect mouth edges, teeth and facial detail in the output.
3. Generate and review
Run the selected task and wait for the queue. Review the result, adjust one setting at a time, and save the output using the controls in the demo.
Common questions
Is the demo free to use?
This public demo runs on a GPU funded by its operator. Wan2.Video does not charge for access. Generation can queue or become unavailable when the service is asleep, busy or no longer funded; this is not a personal ZeroGPU allowance.
Where are my uploads processed?
The selected Hugging Face Space receives your inputs and runs the model. Its maintainer controls processing and retention. The surrounding guide is translated; the demo uses the languages provided by its maintainer.
What if the demo does not load?
Reload the workspace or choose a backup server if available. If your GPU quota is used up, wait for it to reset or try a related tool.