What is this page for?
checking system…

Digital Twin

A voice-cloned, lip-synced version of me — synthesized locally on this Mac with IndexTTS-2 + MuseTalk. Type anything, and "I" say it.

Make me talk

0/400
Ready.

Talking avatar

The video lip-syncs the photo (none yet) to the cloned voice. Drop a new front-facing headshot at digital_twin/input/selfie.jpg to change the face.

About this demo

Your voice was cloned from a short reference recording (IndexTTS-2, zero-shot voice cloning) and the avatar uses MuseTalk for lip-sync — both running locally in MLX on Apple Silicon. Generated audio is cached for a short window then cleaned up. This page is public; generation is rate-limited.
🤖 What this page is for

This page lets you type a line and hear it in Chris's cloned voice, then optionally render a lip-synced video of his photo saying it too. Both run locally on his Mac, so the first try is slow and a video takes a few minutes. Use it to preview how the twin sounds and looks, and to check the models are ready before you try. The assistant beside it explains the wait and helps you phrase a line worth generating.