Tortoise TTS

TorToiSe is a text-to-speech program that prioritizes strong multi-voice capabilities and highly realistic prosody and intonation. It uses both an autoregressive decoder and a diffusion decoder, and can achieve 0.25-0.3 RTF on 4GB VRAM with streaming latency under 500 ms.

AI named Tortoise TTS from December 2025 to September 2026.