# Fish Speech — Multilingual TTS for 80+ Languages

> Fish Speech is a state-of-the-art open-source TTS system supporting 80+ languages. 29K+ GitHub stars. 4B dual-AR model, voice cloning, emotional control with 15K+ tags, real-time inference.

## Install

Save in your project root:

## Quick Use

```bash
# Install
pip install fish-speech

# Or run with Docker
docker pull fishaudio/fish-speech

# Generate speech
fish-speech tts "Hello, this is Fish Speech!" --output hello.wav

# Voice cloning (10-30 second reference audio)
fish-speech tts "Clone my voice" --reference speaker.wav --output cloned.wav
```

---

## Intro

Fish Speech is a state-of-the-art open-source multilingual text-to-speech system with a 4 billion parameter Dual-Autoregressive architecture and reinforcement learning alignment. With 29,000+ GitHub stars, it supports 80+ languages including English, Chinese, Japanese, Korean, Spanish, Arabic, French, German, and more. Fish Speech offers fine-grained emotional control via 15,000+ natural language tags, voice cloning from 10-30 second audio samples, multi-speaker and multi-turn dialogue generation, and real-time inference (0.195 RTF on H200 GPU).

**Best for**: Developers building multilingual voice applications, voice cloning, or emotional speech synthesis
**Works with**: Claude Code, OpenAI Codex, Cursor, Gemini CLI, Windsurf
**Languages**: 80+ (English, Chinese, Japanese, Korean, Spanish, Arabic, French, German, and more)

---

## Key Features

- **80+ languages**: Broad multilingual coverage across global language families
- **4B Dual-AR model**: Natural, realistic speech with sub-word prosody control
- **15,000+ emotion tags**: Fine-grained emotional control via natural language
- **Voice cloning**: Clone any voice from 10-30 seconds of reference audio
- **Multi-speaker dialogue**: Generate conversations with multiple speakers
- **Real-time inference**: 0.195 RTF on H200 GPU, WebUI and API server included

---

### FAQ

**Q: What is Fish Speech?**
A: Fish Speech is an open-source TTS system with 29K+ stars supporting 80+ languages. It uses a 4B parameter Dual-AR model with voice cloning, emotional control via 15K+ tags, and real-time inference.

**Q: How do I install Fish Speech?**
A: Run `pip install fish-speech` or use Docker: `docker pull fishaudio/fish-speech`. Requires a GPU for inference.

---

## Source & Thanks

> Created by [Fish Audio](https://github.com/fishaudio). Research license.
> [fishaudio/fish-speech](https://github.com/fishaudio/fish-speech) — 29,000+ GitHub stars

---
Source: https://tokrepo.com/en/workflows/88c15e9c-439c-4e70-8b8f-cd04efe928c0
Author: AI Open Source