sync.

What is sync.?

sync. is an AI lip-sync platform from Synchronicity Labs that edits the mouth movements of people in existing video so they match a new audio track. Studios, developers and creators use it for visual dubbing, translated content and fixing dialogue after a shoot.

You supply a video and either an audio file or a script for text-to-speech, choose a model, and the system generates a version where the speaker appears to say the new words. The official site lists several models, from the fast lipsync-1.9 to lipsync-2, lipsync-2-pro, the more expressive react-1 and the newer sync-3, which the documentation says supports native 4K. Work can run through a web studio with Lite, Assistant and Advanced modes, or through an API for automated pipelines.

The San Francisco company joined Y Combinator in its Winter 2024 batch, according to its YC profile. Founders include chief scientist Prajwal K R and chief technology officer Rudrabha Mukhopadhyay, both co-authors of the Wav2Lip research paper, alongside chief executive Prady Modukuru and Pavan Reddy. Because the product alters what real people appear to say, you should only process footage of people who have given clear consent.

Key Features

Model choice is the core of sync.: five lip-sync models trade cost against quality, so you can match each job to its budget and screen size.

• Tiered lip-sync models: The range runs from lipsync-1.9 for fast, cheap drafts to lipsync-2-pro and sync-3 for premium detail. Per-second prices on the pricing page rise with each tier, so you can test cheaply and finish on a better model.
• Hard-shot handling: The homepage says the models cope with objects in front of the face, several faces, changing camera angles, low light and fast movement. This matters for real film footage, which rarely offers a clean, front-facing speaker.
• Expressive react-1: The documentation describes react-1 as an expressive model for short clips that carries emotion into the face. It costs more per second than lipsync-2, so it suits short hero shots rather than long talking-head videos.
• Studio with three modes: The web studio offers Lite for quick results, Assistant for a guided workflow and Advanced for detailed control. Editors who never touch code can still upload footage, pick a model and review results.
• Voice clones and text-to-speech: Plans include between 3 and 50 voice clones, and the Creator tier lets you plug in your own TTS API key. You can therefore generate translated speech and matching lips in one workflow.
• Batch API and concurrency: Growth and Scale plans add a Batch API and run 6 or 15 jobs at once. Localisation teams can push whole episode libraries through without queuing each file by hand.

Pros

Research pedigree is why studios trust sync.: its founders co-wrote Wav2Lip, a widely cited lip-sync paper, and the models target difficult production footage.

• Handles real footage: Occlusion, multiple speakers and moving cameras are listed as supported cases. Film and advertising teams can work with existing shots instead of reshooting simple close-ups for dubbing.
• Low entry price: The Hobbyist plan costs $5 a month, and the documentation describes a no-card trial allowance. Freelancers can test the output on their own footage before committing to a larger plan.
• Developer friendly: Official SDKs cover Python 3.8 and later and TypeScript on Node.js 18 and later. Inputs can be public URLs, direct uploads or saved assets, which simplifies pipeline design.
• Fits editing software: Plugins for Adobe Premiere and DaVinci Resolve, plus a ComfyUI node, are documented. Editors can apply lip-sync within the tools where they already cut and grade footage.
• Verification watermark: The site describes a proprietary watermark that lets you check whether a video was modified with sync.. This gives broadcasters and platforms a way to trace synthetic edits.

Cons

Usage-based billing is the main catch with sync., because every second of output is charged on top of the monthly subscription.

• Costs scale with length: sync-3 runs at about $0.13 per second on lower tiers, roughly $8 for each minute of output. Long-form dubbing therefore needs careful budgeting and model choice.
• Short limits on small plans: Hobbyist caps videos at one minute and runs one job at a time. Creators working on full episodes must move to Growth or Scale for 10 or 30 minute limits.
• Watermark on the cheapest tier: The pricing page shows a watermark on Hobbyist output, removed from Creator upwards. Anyone delivering client work should budget at least $19 a month plus usage.
• Broad content licence: The terms grant the company a perpetual, royalty-free licence to user content for running, improving and marketing the service. Studios with confidential footage should negotiate enterprise terms first.
• Misuse and consent risk: Realistic lip-sync can make people appear to say things they never said. Users carry the burden of securing consent and rights, and reputational damage from misuse can be severe.

Who is Using sync.?

Localisation and post-production teams form the main audience, using sync. to make dubbed or rewritten dialogue look natural on screen.

• Dubbing studios: They translate films, series and adverts into other languages. Visual dubbing keeps the actor’s lips in step with the new track.
• Video editors: They fix fluffed lines or script changes after filming using the Premiere or Resolve plugins. This avoids costly reshoots for small dialogue edits.
• Developers and platforms: They build translation, avatar or video apps on the API and SDKs. Batch processing supports large content libraries.
• Content creators: YouTubers and educators localise their own videos for new audiences. Voice clones keep their own voice across languages.

Pricing

sync. charges a monthly subscription from $5 plus per-second generation fees, with a small free allowance noted in its documentation (official pricing page and docs, October 2026).

Usage on Hobbyist and Creator ranges from $0.025 per second for lipsync-1.9 to about $0.167 for react-1, with lipsync-2 at $0.05 and sync-3 at about $0.133.

Disclaimer: per-second rates shift as new models launch, so estimate a real project in the studio calculator first.

What Makes sync. Unique?

Wav2Lip roots set sync. apart: two of its founders co-authored the open research model that shaped modern lip-sync, and the commercial models extend that work to studio footage.

Most avatar tools generate a new presenter from a photo, whereas sync. edits the people already in your video. That makes it a post-production and localisation tool rather than an avatar maker, and the per-model pricing lets teams choose between cheap drafts and 4K finishing passes.

Compatibilities and Integrations

Plugins for Premiere, DaVinci Resolve and ComfyUI bring sync. into existing edit suites, while REST endpoints and SDKs cover custom pipelines.

• API and SDKs: Official libraries support Python 3.8 or later and TypeScript on Node.js 18 or later. Jobs accept URLs, uploads or stored assets.
• Editing plugins: The documentation lists Adobe Premiere, DaVinci Resolve and ComfyUI integrations. Results return into the editor’s own timeline or node graph.
• AI assistants: A ChatGPT plugin and an MCP server are documented. These let assistant tools submit lip-sync jobs on your behalf.
• File formats: Video goes in as MP4 and audio as WAV or MP3. The homepage mentions 4K ProRes output for professional delivery.

sync. Tutorials

A first dubbing test takes six steps in the studio, and the same order applies when you script it through the API.

1. Sign up at sync.so and open the studio.
2. Confirm you have consent and rights for every face and voice in the clip.
3. Upload an MP4 video and a WAV or MP3 audio file, or enter a script for text-to-speech.
4. Choose a model, starting with lipsync-2 for a low-cost test.
5. Generate the clip and review mouth shapes, especially on fast or side-on shots.
6. Re-run key shots on lipsync-2-pro or sync-3, then export or send results to your editor.

How We Rated It

Strong models and wide integrations put sync. at 4.0 out of 5, though per-second costs and broad content licence terms pull the score down.

Summary

sync. suits dubbing studios, editors and developers who need lip movements to match new or translated dialogue in real footage. Its main limitation is cost, since every second of output is billed on top of a subscription and premium models are several times dearer than the basic one. The terms also grant the company a broad licence over uploaded content, which confidential projects should review. Always obtain consent from anyone whose face or voice you alter, and label synthetic edits where appropriate. Test a short, difficult shot on lipsync-2 first, then pay for premium models only where the difference shows.