Clone any voice in 20 seconds locally 🥹
Runs real-time on a CPU in your phone or a Raspberry Pi, a low-end laptop with no GPU😗
Beats ElevenLabs Flash in size & cost, open-source
3× smaller model than ElevenLabs equivalents
221 tokens/sec on mid-range CPU, Real-time inference on CPU only
Instant voice clone from JUST 3 seconds
a foundation model, real-time on CPU laptop phone, Raspberry Pi, no API keys