Tortoise TTS is an open-source multi-voice text-to-speech model known for producing highly expressive and natural-sounding speech with voice cloning capabilities from audio references. It prioritizes audio quality over speed and is used in research and creative audio applications.