OpenSpline: open-source avatar layer adds a realtime lipsynced face to your voice agent in 5 lines
Mikasa0xdev · reddit · 2026-10-15
Developer Mikasa0xdev released OpenSpline, an open-source (MIT) avatar layer for existing voice agents. Supply a portrait and your agent's audio, and it generates synchronized talking video on your own NVIDIA GPU; a lite model runs locally on a single GPU.
Code is on GitHub with a demo at openspline.dev. The project is early-stage but results look promising; the author is soliciting feedback and use-case ideas.
Related event: OpenSpline Open-Sources Real-Time Avatars for Voice Agents(2 posts)→
More from coding & agent
- Benchmarking coding agent harnesses and inference: lessons from $100Ks/month in research spend — abhijithneil · 2026-10-15
- AssetFurnace: Free Open-Source Local Image-to-3D Pipeline for Game Characters — chongdashu · 2026-10-15
- BaoCut's open-source approach to AI bilingual subtitle alignment, in five pitfalls — dotey · 2026-10-15
- Claude Code ships 278 changes in 2 days: 7 updates worth knowing — minchoi · 2026-10-15
- Video edited entirely by Claude Opus 4.5 High using free self-made skills, no editing software — chongdashu · 2026-10-15
- How to Connect Grokbot to Claude Code, Codex or Any AI Agent via HQ Messages — jacob_posel · 2026-10-15