Show HN: A real-time lip-sync avatar so your GPU can sit in standups for you

Category: ai-ml

Tags: lip-sync, ai, video-generation

Score: 6.0/10 (Innovation: 6, Technical: 7, Documentation: 5, Utility: 6)

Wav2Lip is an open-source AI model for lip-syncing videos to arbitrary speech, enabling real-time avatar generation. Its technical approach achieves high accuracy across identities and languages, with a commercial version available via API. The project's combination of research-grade innovation and practical utility makes it interesting for automation and media production.

Target audience: backend devs

Repository: https://heiner-palmen.github.io/project/software-development/automation/ai/2026/05/13/scrumsurvivor.html · Python · 13022 stars

View on Hacker News