SpeakSwap combines multi-speaker AI dubbing with optional, identity-aware lip sync. For an eligible video clip longer than 20 seconds and within your account's dubbing duration limit, it detects speakers and recurring faces, then asks you to confirm who should be lip synced before the final render.
How the lip-sync review works
- Upload a supported video or paste a supported video URL.
- Choose the source and target languages and turn on smart lip sync.
- SpeakSwap prepares the translated dub and detects speakers and recurring faces.
- Confirm or adjust the speaker-to-face matches in the review step.
- Render the lip-synced video, or keep the completed audio dub without face animation.
Lip sync and speaker count are separate controls
Lip sync does not require selecting multiple speakers. Its availability depends on the source being video and the clip duration being longer than 20 seconds and within your account's dubbing duration limit. A solo talking-head clip can use lip sync with the speaker count set to one.
Speaker count serves a different job: it tells the dubbing workflow how many distinct voices to separate for a podcast, interview, panel, course, or co-hosted video. Heavy crosstalk and obscured faces can still require extra review.
Current capability boundaries
The current lip-sync release is intentionally bounded. These limits describe what is available now rather than a promise for every source video.
| Requirement | Current behavior |
|---|---|
| Source | A supported video source is required. Audio-only files can still be dubbed without lip sync. |
| Duration | Longer than 20 seconds and within your account's dubbing duration limit. |
| Speaker count | One or more. Multiple-speaker selection is not required to enable lip sync. |
| Identity matching | The user confirms detected speaker-to-face matches before rendering. |
| No selected face | The completed audio dub remains available without lip-sync animation. |
When multi-speaker dubbing matters
Multi-speaker dubbing matters when the audience needs to follow who is speaking, not only what is being said. Set the speaker count for host-and-guest podcasts, interviews, panels, classes, gaming sessions, and co-hosted creator videos.
Start with the AI dubbing tool, see a multi-speaker dubbing example, or compare the current market in the best AI dubbing tools guide.
FAQ
Does SpeakSwap support smart multi-person lip sync?
Yes. Smart lip sync is optional for eligible video clips longer than 20 seconds and within the account's dubbing duration limit. SpeakSwap detects speakers and recurring faces, then pauses so the user can confirm speaker-to-face matches before rendering.
Does lip sync require selecting multiple speakers?
No. Lip-sync eligibility depends on having a supported video source and an eligible duration, not on selecting two or more speakers. The separate speaker-count control helps the dubbing pipeline handle distinct voices in interviews, podcasts, and panels.
Can I correct the detected speaker and face matches?
Yes. SpeakSwap presents the detected matches for confirmation before the final lip-sync render. If no face is selected for lip sync, the completed audio dub remains available without face animation.
Does lip sync work with audio-only files?
No. Lip sync needs a video source because it analyzes faces and mouth movement. Audio-only dubbing, transcription, text-to-speech, voice cloning, subtitle translation, and stem separation do not require lip sync.