Sync Labs
2 known investors
Sync develops lipsync technology that edits and syncs mouth movements in video to match audio across multiple languages, working on content including movies, podcasts, games, and animation. It offers API access, studio tools, and proprietary watermarking to verify whether a video was modified using its technology.
Also known as sync. Β· sync. labs Β· sync.so Β· Synchronicity Labs Inc.
Investors Β· 2
Also in the syndicate Β· 1
Company profile
researched Aug 2026Sync Labs (branded "sync.", operating as Synchronicity Labs Inc.) develops audio-visual generative models that understand, modify, and synthesize humans in video. Its core capability is lipsync: editing the lip movements of any speaker in an existing video so they match a target audio track, whether that audio is uploaded, cloned, or generated from a script. The company describes its models as working across live-action footage, animation, and AI-generated characters, and in difficult shooting conditions such as objects blocking the face, multiple faces and camera angles, low light, fast camera movement, rapid dialogue, and 4K ProRes material. Models named on the company site include sync-3, lipsync-2-pro, and react-1.
The flagship lipsync-2 model is described as zero-shot video-to-video lipsyncing: it does not require a per-speaker actor, clone, or avatar to be trained or fine-tuned in advance. The company adds a "style preservation" capability, learning how a person speaks from the input video via a spatiotemporal transformer that encodes mouth shapes into a style representation, with a generative transformer synthesizing new mouth movements conditioned on the target speech and that style representation. Supporting components include an active speaker detection pipeline that associates a voice with a face so lipsync is applied only when a given person is speaking, enabling long videos with multiple speakers. The company also offers proprietary watermarking so a video can be uploaded and checked for whether it was modified using sync.'s technology.
Products are delivered through an API, studio tools, and plugins including a Premiere plugin and a ComfyUI node, with tiered workflows described as lite, assistant, and advanced, plus direct collaboration with the company's team for demanding projects. Documented use cases include video translation and dubbing across languages (examples cited include English, Spanish, Hindi, French, German, and Korean), word-level dialogue editing in post-production without reshoots, character re-animation, creative advertising, and AI-generated user-generated content.
Founding story
Sync Labs was founded in 2023 by Rudrabha Mukhopadhyay, Prady Modukuru, Pavan Reddy, and Prajwal K R. Prajwal K R (Chief Scientist) holds a Ph.D. from the University of Oxford under Prof. Andrew Zisserman, and Rudrabha Mukhopadhyay (CTO) completed a Ph.D. at IIIT Hyderabad in audio-visual deep learning; both were primary authors of Wav2Lip, the open-source lipsync model published in 2020 that the company cites as the most prolific to date. CEO Prady Modukuru previously worked on AI-powered cybersecurity products at Microsoft. Pavan Reddy, an IIT Madras alumnus and repeat venture-backed founder, leads sales, operations, finance, and strategic roadmap.
Business model
Sync Labs sells access to its lipsyncing models as developer infrastructure and tooling: a hosted API, studio tools, and editing plugins, with published pricing plans and a sales-led track for larger customers. Its stated approach is to train state-of-the-art generative models, productize and host them at scale, grow virally through content, and upsell into enterprise. A sales inquiry form segments prospects by company size (0-10 through 1000+ employees) and expected monthly usage (under 60 minutes, 1-10 hours, 10+ hours), indicating usage-based commercial packaging.
Paid plans for API and product access (a "scale plan" plus credits are referenced in a Y Combinator promotional deal offering four months of the scale plan free plus $1,000 in credits), alongside enterprise/direct engagements handled by the sales team.
Traction
The Y Combinator launch states thousands of developers and businesses integrate the company's APIs to build generative video workflows, including video translation, word-level video editing, and character re-animation. The founding team's earlier open-source Wav2Lip model has over 10,000 GitHub stars. Team size is reported at 18.
βΈFull profile β market position, technology, go-to-market, geography, history, risks & controversies
Market position
Positioned as an AI lipsync and visual dubbing infrastructure provider serving thousands of developers and businesses that build video translation, dialogue editing, and character re-animation workflows. It is backed by Y Combinator (Winter 2024), Founders, Inc., and a seed round led by GV.
The company positions itself on zero-shot operation combined with style preservation β editing an arbitrary speaker without training a per-person model while retaining their speaking style, skin texture, and other identity traits, which it states was not achieved by prior zero-shot lipsync models. Other cited differentiators are cross-domain coverage (live-action, animated, and AI-generated humans), robustness in difficult shooting conditions, multi-speaker handling via active speaker detection, provenance watermarking and verification, and research lineage as the original authors of Wav2Lip.
Technology
Zero-shot video-to-video lipsync models. lipsync-2 uses a spatiotemporal transformer to encode the mouth shapes observed in an input video into a speaker style representation, and a generative transformer that synthesizes new mouth movements conditioned on target speech plus that style representation, preserving speaker-specific traits without fine-tuning. A state-of-the-art active speaker detection pipeline pairs each voice with the correct face so edits are applied only to the actively speaking person, supporting long multi-speaker videos. Additional models listed are sync-3, lipsync-2-pro, and react-1. The company also operates proprietary watermarking that allows a video to be verified as having been modified with its technology. The founding team authored the open-source Wav2Lip lipsync model (published 2020, 10,000+ GitHub stars).
Go-to-market
Self-serve product and API with public documentation, changelog, and pricing; distribution through ecosystem plugins (Adobe Premiere, ComfyUI); community channels on Discord, X/Twitter, Instagram, and YouTube; content-driven and viral growth followed by enterprise upsell; a private beta and demo-request pipeline into a sales team; and Y Combinator-linked promotional offers.
Developers and businesses embedding generative video workflows into their own products, video content creators, and enterprises; the company also works directly with studios and teams on projects requiring exacting quality. Its intake form addresses companies from under 10 to over 1,000 employees.
Geography
Headquartered in San Francisco, with founders and research roots connected to the University of Oxford and IIIT Hyderabad; the product supports multilingual dubbing including English, Spanish, Hindi, French, German, and Korean.
History
Founded in 2023 by the team behind the open-source Wav2Lip project, the company participated in Y Combinator's Winter 2024 batch and is based in San Francisco. It raised a $5.5M seed round led by GV and is a portfolio company of Founders, Inc. It subsequently launched lipsync-2, described as a new generation of zero-shot lipsync models with style preservation, and its site lists later models including sync-3 (promoted as its most intelligent lipsyncing model), lipsync-2-pro, and react-1, along with API, studio tools, Premiere and ComfyUI integrations, and watermark verification. The corporate entity is identified as Synchronicity Labs Inc.
Risks & controversies
The technology modifies the appearance of real people's speech in video, which the company addresses through proprietary watermarking that lets any video be checked for whether it was altered using its models.
Compiled by commissioned research from 8 cited public sources β announcements, filings, and press listed under research sources below.
Key figures
latest reportedCompany-reported or press-reported figures, each dated to when it was claimed β not independently audited.
Timeline Β· 4
launches, deals, and filingssync. offers proprietary watermarking technology allowing users to upload a video and verify whether it was modified using the company's technology.
The company's site promotes sync-3, described as its most intelligent lipsyncing model, alongside lipsync-2-pro and react-1.
The company launched lipsync-2, described as a zero-shot lipsyncing model that edits any speaker's lip movements to match any audio without per-speaker training, introducing style preservation and working across live-action, animated, and AI-generated humans. The launch was paired with a YC deal offering four months of the scale plan free plus $1,000 in credits.
sync. is listed as a Y Combinator Winter 2024 batch company, based in San Francisco, with active status.
Dated company events from announcements, filings, and press; legal rows summarize public dockets and regulator releases.
βΈResearch sources Β· 8
primary sources listed
- Sync Labssynclabs.so Β· web
8 public sources were cited for this profile; the first-party ones are listed here.
Frequently asked questions
- What does Sync Labs do?
- Sync Labs builds zero-shot AI lipsync models and an API that edit mouth movements in video to match any audio.
- Who are Sync Labs's investors?
- Sync Labs's investors include Creator Ventures.