1. Accurate Tone Color Cloning. OpenVoice can accurately clone the reference tone color and generate speech in multiple languages and accents.
Global
IndexTTS is a zero-shot text-to-speech system that clones a voice from a single reference audio clip.
5B CosyVoice-300M CosyVoice-300M-SFT CosyVoice-300M-Instruct model and CosyVoice-ttsfrd resource.
#6321 · +1 · 1d ·GitHub
FireRedTTS3 is a unified speech generation and editing system built on semantically enriched continuous speech representations. It comes in two variants:
AuK: An Open-Source Foundational Model for Speech Generation and Editing