Lip sync a video to new audio in Claude and ChatGPT

Yes. Connect Makemancer and Claude, ChatGPT or Codex can lip sync a video to new audio: give a clip and a voice, and Sync 3 or PixVerse redraws the speaker's mouth to match while the rest of the picture stays as it was. A 10-second clip costs at most $1.40 with PixVerse or $4.20 with Sync 3.

Bring
A video with a visible face and a voice
Get
The same video, speaking the new words
From
80 credits · $0.80

Made with it

Our own results, first takes. Each caption says what it took.

A new line for an existing clip

A 6-second clip of a fictional woman talking, and a new line we generated as speech. Sync 3 replaced the clip's sound with the line and redrew her mouth to say it; her face, the room and the camera are the original frames. Turn the sound on to compare.

InA fictional woman recording a video message, the source clipMiniMax H3 · 768p · 6 s · 16:9 · 96 credits
InThe new line for the lip sync, Gemini TTSHi everyone, quick update: the new studio opens next Monday, and you are all invited.Gemini 3.8 Flash TTS · 2 credits
ResultThe clip lip-synced to the new line by Sync 3Sync 3 Lip Sync · 295 credits

The voice is already on the video

Here the line first replaced the clip's own sound, starting a second in, so the words and the mouth did not match. Given only the video, Sync 3 found the voice in it and fixed the mouth. Captions show the quote; the final charge counts only the seconds with speech, so it can be lower.

InThe source clip with the new line laid on its own soundtrackAdd Audio To Video · 2 credits
ResultSync 3 lip-syncs the clip to the voice already in itSync 3 Lip Sync · 297 credits

How to do it

  1. Connect Makemancer

    Add Makemancer to Claude, ChatGPT, Claude Code or Codex. The steps for each client are on the connection page.

  2. Give the video and the voice

    Attach or link the video, then the voice: a recording, a line you generate as a voiceover, or speech in your cloned voice. Leave the voice out if it is already on the video. If several people are on screen, say who speaks.

    Lip sync this video to this audio with Sync 3. The woman in the denim shirt is the one speaking. Tell me the cost first.
  3. Check the mouth, then the timing

    Watch it once with sound. If a line belongs at a set moment, give the second it starts and only that stretch is synced, with the clip's own sound around it.

    The new line should start 1.5 seconds in, not at the start. Redo it with the voice starting at 1.5 s and keep the original sound before it.

Which model to use

ModelBest forPriced forCredits
PixVerse Lip Synclip-sync/pixverseThe default and the cheaper one: finds the speaker from the picture.Opens after your first credit purchase12 s video164$1.64
Sync 3 Lip Synclip-sync/sync-3The premium one, and the one where you can say who speaks when several people are on screen.Opens after your first credit purchase12 s video492$4.92
H3 Max Lip Syncvideo/minimax-h3-max-lip-syncNo video yet? Animates one still portrait to a recording, as long as the audio runs.8 s audio input80$0.80

What lip sync changes, and what it keeps

Only the mouth and the area around it are redrawn, and only during spoken lines, each with a short margin. Everything else is the original footage: the rest of the face, the background, the motion and the cut. The voice you give replaces the video's sound; to keep the original sound under a short line, give the second the line starts and only that stretch changes.

Sync 3 or PixVerse

PixVerse is the default and the cheaper choice for a single clear speaker. Sync 3 costs more, and it is the one to use when several faces are on screen, because you can name who says each line; both runs above used it. Both take videos up to 15 minutes. Lip sync opens after your first credit purchase.

Starting from a photo instead

If you only have a portrait, there is no video to sync yet. A talking avatar turns one photo and a voice into a speaking clip, with lip movement made from the audio. To put a different person into a video you already have, see character swap.

Questions

How do I lip sync a video to new audio?

Connect Makemancer, give the assistant the video and the new voice, and ask for a lip sync. It quotes the cost, replaces the video's sound with the voice and redraws the mouth to match. The result is a new video file with a permanent link.

Does it work with my own voice?

Yes. Use a recording of yourself, or clone your voice once and generate any line in it, then lip sync a video to that line. Only use footage of people who agreed to appear in it.

Can it sync to the voice already on the video?

Yes. Leave the new audio out and the model syncs the mouth to the speech in the video's own soundtrack, as in the second run above.

What if two people are talking?

Use Sync 3 and say who says what, such as the man on the right or the first voice is the owl. Lines whose speaker is off screen are left alone and not charged.

Which videos work best?

A clear, mostly front-facing human face that stays in view, up to 15 minutes long. Animals, cartoons and faces turned away can fail or come out distorted; the assistant says so before a paid attempt.

What does it cost?

Lip sync is charged per second of voice, up to the video's length: a 10-second clip costs at most $1.40 with PixVerse or $4.20 with Sync 3. Only the seconds with speech are charged in the end, so it is often lower. Lip sync opens after your first credit purchase.