sync. labs Information

sync. is a research-driven production company built by scientists, artists, and engineers advancing the future of AI video dubbing, video localization, and AI-powered video editing. Our research focuses on generative facial performance: teaching models to understand and manipulate how people look, move, and express themselves on screen. Our journey began in 2019, when our founders built the world’s first zero-shot AI lip sync model capable of syncing lips to any audio on any video without per-speaker training or fine-tuning - Wav2Lip. That breakthrough laid the foundation for the tools we build today, spanning AI video translation, visual dubbing, dialogue replacement, and lip sync API infrastructure for studios and developers. At sync., research and production operate in a tight feedback loop. We improve our models by working on real titles, real localization workflows, and real production constraints, not just benchmarks. What we ship this week informs what the model learns next. Backed by that foundation, we’re building the infrastructure for the next generation of AI video editing, AI dubbing platforms, and global content workflows. Y Combinator W24.