What video formats are supported as input?

MP4 and MOV at common resolutions (720p, 1080p, 4K) and frame rates (24/25/30/60 fps). The video should have a clearly visible face in roughly two-thirds or more of the frame for the sync to lock on.

How long can the input video be?

Most jobs run on clips up to 5 minutes. Longer-form content (10+ minute explainers) is supported on higher-tier plans where compute budget allows. Splitting a long video into scenes is usually a better pattern than running one 20-minute job.

Does it work for non-English audio?

Yes, the lip-sync model is phoneme-driven, not language-driven. It handles English, Spanish, Portuguese, French, German, Italian, Hindi, Mandarin, Japanese, Korean, Arabic — and the 35-language set AdLit's TTS supports. Tonal languages (Mandarin, Vietnamese) work too.

How accurate is the lip-sync?

Tight enough to fool casual viewers. Close-up extended shots (10+ seconds of the same face on screen) are the hardest case — for those, you'll occasionally see a syllable that's slightly off. For typical UGC pacing (cuts every 3–6 seconds), the sync is effectively perfect.

NewStudio 2.0 — Talking Actors at 1080p

Perfect Sync with AI Lip Sync
in Any Language

Generate perfectly lip-synced videos instantly.
Seamless localization for global content.

Generate Your Ad

Continue with Google

Talking Actors 650+ AI Avatars Lip Sync Voice Cloning Text-to-Speech Product Videos Shorts Generator AI UGC Generator Talking Actors 650+ AI Avatars Lip Sync Voice Cloning Text-to-Speech Product Videos Shorts Generator AI UGC Generator

Batch Generation 35-Language Dubbing Background Remover Auto Captions Character Replace B-Roll Generator Hooks & Gestures Voice Library Batch Generation 35-Language Dubbing Background Remover Auto Captions Character Replace B-Roll Generator Hooks & Gestures Voice Library

Ad Video Generator Facebook Ad Generator TikTok Ad Generator AI Ads Video API Lip Sync API Affiliate Tools Content Generator Ad Video Generator Facebook Ad Generator TikTok Ad Generator AI Ads Video API Lip Sync API Affiliate Tools Content Generator

AdLit's lip-sync engine takes an existing video and a new audio track and re-animates the mouth to match. Same body language, same head movements, same lighting — different words coming out. It's the cleanest way to localize a piece of UGC into another language, replace a voiceover after launch, or fix muffled production audio without re-shooting.

The reality

Why this is hard today

Localization re-shoots are brutal

Filming the same ad in five languages means flying the creator out five times, or hiring five creators and losing brand consistency. Lip-sync removes the re-shoot entirely.

Post-launch script changes

Legal flags one phrase the day after the ad goes live. Without lip-sync, that's a full re-render or a re-shoot. With it, you swap the audio and the mouth follows.

Bad production audio kills good footage

Great visual take, terrible audio (wind, lapel mic failure, room noise). Re-recording the VO and lip-syncing it onto the original take saves the shoot.

What you can ship

Built for these workflows

Dub one ad into 15 languages

Generate the ad once with your hero creator. Render lip-sync passes with localized voiceovers for each market. Same person, same emotion, fifteen languages.

Fix audio after launch

Found a typo in the script the day after launch? Render a new audio track, lip-sync it onto the existing video, and replace the asset without re-rendering everything else.

Save bad-audio footage

Upload a great take with muffled audio + a clean studio re-record of the script. AdLit aligns the new audio to the original mouth shapes, salvaging the visual.