Lux APIs

Lux builds fast voice-cloning speech models. Its TTS model is distilled to four inference steps, cloning a voice from a short reference clip at 48 kHz.