Kyutai, an open AI research lab, has released a groundbreaking streaming Text-to-Speech (TTS) model with ~2 billion parameters. Designed for real-time responsiveness, this model delivers ultra-low ...