Turn a portrait and a script into a talking avatar. Three lip-sync engines: an instant in-browser preview, MuseTalk for photoreal rendering on your own machine (Apple MPS/CUDA), and HeyGen v3 for cloud renders that also move the head. TTS via Gemini, ElevenLabs (with voice cloning from a video clip), OpenAI or Piper. FastAPI backend, no build step.
-
Updated
Jul 12, 2026 - JavaScript