Fish Speech
SOTA open-source TTS with an excellent community reputation
A SOTA open-source TTS with an excellent community reputation.
Text-to-speech, speech-to-text, and voice cloning, open source.
Key features
- Multilingual TTS trained on 10M+ hours across 80+ languages
- Dual-AR architecture with RL alignment for natural speech
- Sub-word prosody and emotion control via natural-language tags
- Multi-speaker and multi-turn conversation generation
Good for
- Natural multilingual voice generation for applications
- Emotionally expressive narration and dialogue
Stack & tags
ttstext-to-speechmultilingualvoice-generationemotion-controlopen-weightsprosody
Screenshots
Screenshots from the project’s GitHub README.
Similar open-source projects
Whisper
★ 109,795OpenAI's open-source speech recognition model — the benchmark for multilingual STT
GPT-SoVITS
★ 62,271TTS/cloning trainable on 1 minute of audio — a legendary Chinese-community project