A lip sync and visual dubbing platform that re-syncs any face to new audio in any language.
Last verified · 2026-06-23 · by Moe Ameen
sync. (at sync.so, from sync.labs) is an AI platform built around one hard problem: making a face's mouth match a new audio track convincingly. You upload a video, give it a voice — clone an existing one, generate a new one, or keep the audio already in the clip — and sync. re-renders the lips so the speaker looks like they actually said the new words. The same technology powers visual dubbing: feed a translated voiceover and the on-screen presenter's mouth moves to the new language instead of looking like an obvious overdub.
The company came out of Y Combinator (W24) and was founded by the team behind Wav2Lip, the widely used open-source lip-sync project. Its current flagship is sync-3, which the company positions as its most capable model: rather than stitching together isolated frames, it builds an understanding of the person across the whole shot and generates the frames together, which is what lets it hold up through close-ups, occlusions, awkward angles, and low light while preserving the original performance. Earlier models including lipsync-2 (a zero-shot model that needs no per-speaker training) and lipsync-2-pro remain available, and a react-1 model targets reactive, expression-driven output.
sync. is API-first as much as it is an app. You can work in its web Lipsync Studio, call the models through an API and SDKs, or run them inside editors via plugins for Adobe Premiere Pro and ComfyUI. It also ships provenance tooling — a watermark scheme meant to let people verify whether a video was modified with sync.
A few things to be clear about. sync. is a lip sync and dubbing engine, not a text-to-video generator: it re-animates a mouth on footage you already have, it does not invent a scene from a prompt. It is also not a publisher — it returns a synced MP4, not a finished, captioned, scheduled post. Model names, language coverage, and resolution ceilings change as the platform ships, so treat any specific spec as a snapshot and check the docs for current limits.
sync. hands you one thing: a clip where the mouth is finally right. That is the hard, specialized part — and it is also where most people stop, because a single synced MP4 is not a content schedule. Kompozy is the layer that turns that one clean clip into a week of published posts. Drop a sync. export into Kompozy and it burns in branded, on-style captions, reframes the clip to each platform's aspect ratio, and lets you stack hook text and overlays through HyperFrames so the silent-autoplay first second lands. Then it schedules and publishes the same clip across the nine supported platforms — TikTok, Instagram Reels, YouTube Shorts, Facebook, LinkedIn, X, Pinterest, Threads — in one pass instead of you re-uploading into six apps by hand.
The fan-out is where the pairing earns its keep. A single sync.-dubbed video can seed a whole content unit in Kompozy: the video itself for short-form feeds, plus a quote card pulled from the script, a text post and a thread written in your own voice through your Persona Brief, and a blog or newsletter draft off the same transcript. So one re-synced clip becomes a coordinated set across video, image, and text rather than a single upload. And when you need net-new talking-head video that sync. can't make from scratch — there is no source footage yet — Kompozy generates it directly with its Persona Shorts and Persona HeyGen avatar formats. sync. perfects the mouth on footage you have; Kompozy generates what you don't and ships all of it.
sync. is an AI lip sync and visual dubbing platform from sync.labs (YC W24), founded by the team behind the open-source Wav2Lip project. You upload a video and a voice, and it re-renders the speaker's mouth to match the new audio — useful for dubbing, re-recorded audio, and localizing one video into multiple languages.
sync. has a free way to try it, then paid plans that, at the time of writing, started at $5/month (Hobbyist) and rose through Creator ($19/mo), Growth ($49/mo), and Scale ($249/mo), each with a per-second usage charge on top, plus custom Enterprise. Watermark-free output and active speaker detection start on the Creator tier. Check sync.so/pricing for current numbers.
lipsync-2 is a zero-shot model that syncs a speaker without per-speaker training. sync-3 is the newer flagship that builds a whole-shot understanding of the person and generates frames together rather than stitching them, which helps it hold up through close-ups, occlusions, hard angles, and low light. Both are available; sync-3 is positioned as the most capable.
No. sync. re-animates the mouth on video you already have to match new audio — it is a lip sync and dubbing tool, not a text-to-video generator. If you need talking-head video created from scratch, that is a different category; Kompozy generates it directly with its Persona Shorts and Persona HeyGen avatar formats.
sync. returns the synced MP4 but does not publish it. Bring the export into Kompozy to add branded captions, reframe it per platform, and schedule and publish across TikTok, Reels, YouTube Shorts, X, LinkedIn, and more from one queue — and fan the same clip out into supporting posts in your voice.