Skip to content

Updated

Lip-Sync Dubbing

AI Lip Sync Dubbing Voice and Lips in a New Language

AI lip sync dubbing translates the speech in a video, re-voices it in another language and adjusts the speaker's mouth movements to match the new audio. With AutoDub you upload a video or paste a YouTube, TikTok or Instagram link, pick a language, and get the dub in a clone of the speaker's own voice, lip-synced, with subtitles.
Lip-sync in every dubThe speaker's cloned voiceSRT and VTT subtitles70+ languages
Try Lip-Sync Dubbing FreeView Pricing
English

How it works

How to Lip-Sync Dub a Video

Four steps from your video to a dub whose lips match the new language

  1. Step 1: Upload or paste a link

    Upload an MP4, MOV, M4V or WebM file up to 2 GB and 30 minutes, or paste a public YouTube, TikTok or Instagram link.

  2. Step 2: Pick the new language

    Choose one of 70+ languages. Each dub is one language; on a paid plan you can pick up to five at once, and each becomes its own dub.

  3. Step 3: AutoDub dubs and lip-syncs

    The speech is transcribed, translated and re-voiced in a clone of the speaker's voice, then the mouth movements are matched to the new audio. Most dubs are ready in 5–15 minutes.

  4. Step 4: Watch it, then download

    Watch the dub in your dashboard and check the close-ups; download the MP4 and the SRT/VTT subtitles with a plan or a one-time unlock.

How AutoDub Matches the Lips to the New Language

Every AutoDub dub goes through the same steps: speech recognition, translation, a voice clone built from the speaker's own audio, and lip-sync. Lip-sync is not an extra you switch on or pay for separately. It is part of every dub, on the free trial and on every plan.

The lip-sync step adjusts the speaker's mouth movements to the translated audio, so a person talking to the camera looks as if they said the new words. A sentence rarely takes exactly as long in one language as in another, so the timing of the dub can stretch a little where the new language needs more time.

Your voice stays yours: the new language is spoken by a clone of the person in the video, not by a stock narrator. Each dub also comes with subtitles in the new language as SRT and VTT files, which you can read in the player and download with the video.

Demo

Watch One Video Dubbed into Three Languages

Press play, then switch languages and judge the voice and the speaker's mouth for yourself.

03 languages · one source

  1. EnglishEN
  2. EspañolESSpanish
  3. FrançaisFRFrench

ENEnglish

ESSpanishEspañol

FRFrenchFrançais

Same video, dubbed into Spanish and French with AI voice cloning

Comparison

Lip-Sync Dubbing, Voice-Over or Subtitles?

Three ways to put a video into another language, and when each one fits

FeatureSubtitles onlyVoice-overLip-sync dub
What viewers hearThe original audioA translated voice over the pictureThe translation, in the speaker's cloned voice
Mouth matches the new wordsNoNoYes
Viewers have to readYesNoNo
Fits bestViewers who prefer the original soundNarration, slides and screen recordingsTalking-head videos, courses, interviews and ads
With AutoDubSRT and VTT files come with every dubNot offered on its ownEvery dub, trial included

With AutoDub you never choose between them: each dub is lip-synced and comes with its subtitle files.

Where Lip-Sync Dubbing Works Best

Lip-sync needs a mouth it can see. The most natural results come from a face turned towards the camera, in good light, with one person speaking at a time. Side profiles, a hand or microphone in front of the mouth, fast movement, a face that is small in the frame, or two people talking over each other are where a lip-synced dub can look less exact. When no face is on screen, as in a screen recording or a slideshow, there are no lips to match, so the cloned voice and the subtitles are what your viewers notice.

Matching lips to new speech is a hard research problem. Researchers who presented the Wav2Lip model at ACM Multimedia in 2020 wrote that earlier systems were accurate on static images or on people seen during training, but “fail to accurately morph the lip movements of arbitrary identities in dynamic, unconstrained talking face videos, resulting in significant parts of the video being out-of-sync with the new audio.” That is why a clear, steady view of the speaker still matters today.

Source: K R Prajwal, Rudrabha Mukhopadhyay, Vinay Namboodiri and C V Jawahar, “A Lip Sync Expert Is All You Need for Speech to Lip Generation In The Wild”, ACM Multimedia 2020, arxiv.org/abs/2008.10010, checked 4 November 2026.

Viewers are also getting used to dubbed video that looks right. On 4 February 2026 YouTube wrote that it was “currently testing a Lip Sync pilot that subtly matches the speaker’s lip movements to the translated audio so a dubbed video feels as seamless as watching the original”, and that in December it “averaged more than 6 million daily viewers who watched at least 10 minutes of auto dubbed content.”

Source: YouTube Official Blog, “Unlocking a global audience with auto dubbing” (4 February 2026), blog.youtube/news-and-events/youtube-auto-dubbing-expressive-speech, checked 4 November 2026.

Comparison

Lip-Sync Dubbing Limits Free Trial or a Plan

What one lip-synced dub can be, before you pay and with a plan

FeatureFree trialWith a plan
VideosOne video, one languageAs many as your minutes cover
Length per videoUp to 2 minutesUp to 30 minutes
Upload or linkMP4, MOV, M4V and WebM up to 2 GB, or a YouTube, TikTok or Instagram linkThe same
LanguagesOne language per dubOne language per dub; up to five per submission, each its own dub
Lip-sync and voice cloneYesYes
Watch the dub in your dashboardYesYes
TurnaroundMost videos in 5–15 minutesThe same
Download the MP4 and SRT/VTT subtitlesWith a one-time unlock ($4.99)Yes
PriceFreeFrom $39 a month; yearly plans 20% off

Subscribers can top up with minute packs from $25. The free trial needs a verified e-mail address.

FAQ

AI Lip Sync Dubbing FAQ

Faces, speakers, timing and checking the result

6 answers

Yes. Lip-sync is part of every AutoDub dub, including the free trial; there is nothing to switch on and no separate price. Each dub is translated, spoken in a clone of the speaker's voice and lip-synced to the new audio, with SRT and VTT subtitles.

AutoDub does not refuse a video because of who is in it. How natural the lips look depends on how clearly the mouth can be seen: a face towards the camera in good light gives the best result. Glasses sit above the mouth; a thick beard, a mask, a hand or a microphone in front of it, a side profile or a very small face can make the lip-sync less exact. Try your own footage with the free trial first.

Yes, a video with more than one person can be dubbed, and it works best when people take turns. When two people talk over each other, keeping the voices apart and matching each mouth is where mistakes appear, and AutoDub has no setting for the number of speakers. Watch the dub before you publish it: AutoDub has no script editor, so if a dub is not right, improve the source (clearer audio, one speaker at a time) and dub it again, which uses minutes again.

At AutoDub there is no higher price: lip-sync and the cloned voice are part of every plan minute, and of the free trial, so the only question is whether your video shows a speaking face. If it does, lip-sync is what makes the dub look recorded in the new language. If it is narration over slides or a screen recording, the voice alone carries it and the lip-sync simply has nothing to match. Plans start at $39 a month.

The voice in the dub is a clone of yours, so viewers still hear you, now in the new language. In the picture, the mouth movements are adjusted to the translated audio, and the timing can stretch a little where the new language needs more time. Your original video stays in your account until you delete it.

Use lip-sync dubbing when the speaker's face is on screen: viewers notice when a mouth does not match the words. A voice-over is enough for narration over slides, b-roll or screen recordings. With AutoDub you do not have to choose, because every dub is lip-synced wherever a speaking face is visible.

Specs and limits

What AutoDub does not do yet.

Worth knowing before you dub a video:

  • No script editor. You cannot edit the transcript or the translation of a dub, so watch every dub before you publish it.
  • No public API. Every dub is started in the web app; there is no API for teams yet.
  • One language per dub. Each dub has one target language and is billed as its own dub. Paid plans can pick several languages in one submission; the free trial is one video in one language.
  • 30 minutes and 2 GB per video. A longer or larger video has to be split into parts before you upload it.
  • Subtitles only in the new language. The SRT and VTT files are in the dubbed language; there is no subtitle file in the original language.
  • Links from three sites. AutoDub fetches YouTube, TikTok and Instagram links; a video from anywhere else has to be uploaded as a file.
  • No accent or dialect setting. You pick a language, such as Spanish, not a regional variety of it.
  • No automatic deletion. Nothing is removed for you: your videos and dubs stay in your account until you delete them in your dashboard.
  • Downloads need a plan or an unlock. A free trial dub can be watched, not downloaded: the video and its subtitles download with a plan or a one-time unlock from $4.99.

Last verified · Full documentation

Next steps

Related Pages

Try AI Lip Sync Dubbing on Your Own Video

Your first video, up to 2 minutes, is free to try and watch.

Start Free TrialView Plans