Audio to video lip sync
Bring the new audio. Keep the visible speaker.
Use this route when the replacement speech already exists. Upload the finished audio you are allowed to use, pair it with a clear single-speaker video, and review the matched result before publishing.
Start with this inputNeed another route? Compare all three speech inputs.
This approved A/B example shows the kind of short, single-speaker update Lip Sync is designed to make. Review every generated result before publishing.
Create
Bring one clear video. The matching speech mode is already selected.
You can switch modes at any time. The same price, consent, queue and result rules apply across all routes.
What should they say?
Input guide
Give the audio a clean signal
Export the final audio
Use the approved MP3, WAV or M4A rather than a rough mix with music or overlapping speakers.
Upload the matching video
Choose a clip with one clear, evenly lit speaker whose mouth remains visible.
Generate and review
Check timing and visible details in the returned MP4 before you use it in an edit or publish it.
- MP3, WAV and M4A are supported in the starter workflow, up to 8 seconds of speech.
- Speech without music, overlap or heavy background noise gives the mouth-matching step a clearer signal.
- The source video should be at least as long as the new speech and show one visible face.
Questions
Audio to video lip sync questions
Can I upload music or a podcast with several people?
No. This route is designed for short, clean single-speaker speech. Music and overlapping voices make matching less reliable.
What if my video is shorter than the audio?
Use a source clip at least as long as the speech. Shorter clips may need to loop, so inspect the final result carefully.
Can I record instead of uploading a file?
Yes. Switch to Record audio in the shared tool if your browser microphone is available.