[](https://github.com/Akshay090/svg-banners)
## 👉🏻 CosyVoice 👈🏻
**CosyVoice 3.0**: [Demos](https://funaudiollm.github.io/cosyvoice3/); [Paper](https://arxiv.org/abs/2505.17589); [Modelscope](https://www.modelscope.cn/studios/FunAudioLLM/Fun-CosyVoice3-0.5B); [CV3-Eval](https://github.com/FunAudioLLM/CV3-Eval)
**CosyVoice 2.0**: [Demos](https://funaudiollm.github.io/cosyvoice2/); [Paper](https://arxiv.org/abs/2412.10117); [Modelscope](https://www.modelscope.cn/studios/iic/CosyVoice2-0.5B); [HuggingFace](https://huggingface.co/spaces/FunAudioLLM/CosyVoice2-0.5B)
**CosyVoice 1.0**: [Demos](https://fun-audio-llm.github.io); [Paper](https://funaudiollm.github.io/pdf/CosyVoice_v1.pdf); [Modelscope](https://www.modelscope.cn/studios/iic/CosyVoice-300M)
## Highlight🔥
**CosyVoice 3.0** is here! Building on the strengths of version 2.0, it achieves a leap in "in-the-wild speech generation" with larger scale, more powerful technology, and broader coverage—delivering more natural, accurate, and versatile speech synthesis capabilities.
### Multilingual & Wide Coverage
- **Expanded Language** Support: Covers 9 universal languages (Chinese, English, Japanese, Korean, German, Spanish, French, Italian, Russian) and 18+ Chinese dialects/accents, supporting zero-shot multilingual/cross-lingual voice cloning.
- **Massive Data Upgrade**: Training data scaled from ten thousand hours to 1 million hours，enhancing cross-lingual generalization and synthesis stability.
- **Polyglot Transformation**: Enables monolingual speakers to generate speech in multiple languages, with WER/CER below 4% for most languages.
### High Precision & Robustness
- **Superior Content Consistency**: Compared to 2.0, Chinese test set CER relative improves 44%, English test set 51%, and complex scenario (test-hard) CER drops from 6.83% to 5.09%.
- **Pronunciation Perfection**: Features pronunciation inpainting technology with high correction rate for polyphonic characters and rare words, solving long-standing pronunciation challenges.
- **Strong Scene Adaptability**: Excels in noisy backgrounds, tongue twisters, and domain-specific terms; self-training text normalization supports raw text direct synthesis, compatible with special symbols.
### Natural & Expressive Experience
- **Rich Style Control**: Supports 100+ speaking styles (emotions, speeds, roles, accents) and fine-grained instructions ([laughter], [breath], emphasis).
- **Advanced Emotion CloningControl**:: Optimized with DiffRO-EMO, leading in emotion accuracy for both text-related and text-unrelated scenarios.
### Innovative Core Technology
- **DiffRO Optimization**: Innovative reinforcement learning method, directly optimizing speech tokens, improving WER by 20%-50% in low-resource/cross-lingual scenarios, applicable to other LLM-based TTS models.
- **Supervised Multi-task Tokenizer**: Trained on ASR, SER, LID, etc., better capturing paralinguistic information like emotion and pronunciation style.
- **Pronunciation & Text Normalization**: Supports word-phoneme mixed sequence modeling; LLM-based self-training TN eliminates manual rules.
### Real-time & Flexible Deployment
- **Multi-mode Support**: Single model compatible with offline mode, one-way streaming, and bidirectional streaming, meeting diverse deployment needs.
- **Ultra-low Latency**: Achieves latency as low as 150ms while maintaining high-quality audio output.
- **Hot Fix Capability**: Supports SSML hot fix and built-in TN capability, enhancing usability and adaptability to complex text formats.
## Roadmap
- [x] 2025/12
- [x] release CosyVoice3-0.5B base model and its training/inference script
- [x] release CosyVoice3-0.5B modelscope gradio space
- [x] 2025/08
- [x] Thanks to the contribution from NVIDIA Yuekai Zhang, add triton trtllm runtime support and cosyvoice2 grpo training support
- [x] 2025/07
- [x] release CosyVoice 3.0 eval set
- [x] 2025/05
- [x] add CosyVoice2-0.5B vllm support
- [x] 2024/12
- [x] 25hz CosyVoice2-0.5B released
- [x] 2024/09
- [x] 25hz CosyVoice-300M base model
- [x] 25hz CosyVoice-300M voice conversion function
- [x] 2024/08
- [x] Repetition Aware Sampling(RAS) inference for llm stability
- [x] Streaming inference mode support, including kv cache and sdpa for rtf optimization
- [x] 2024/07
- [x] Flow matching training support
- [x] WeTextProcessing support when ttsfrd is not available
- [x] Fastapi server and client
## Install
### Clone and install
- Clone the repo
``` sh
git clone --recursive https://github.com/FunAudioLLM/CosyVoice.git
# If you failed to clone the submodule due to network failures, please run the following command until success
cd CosyVoice
git submodule update --init --recursive
```
- Install Conda: please see https://docs.conda.io/en/latest/miniconda.html
- Create Conda env:
``` sh
conda create -n cosyvoice -y python=3.10
conda activate cosyvoice
pip install -r requirements.txt -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host=mirrors.aliyun.com
# If you encounter sox compatibility issues
# ubuntu
sudo apt-get install sox libsox-dev
# centos
sudo yum install sox sox-devel
```
### Model download
We strongly recommend that you download our pretrained `CosyVoice2-0.5B` `CosyVoice-300M` `CosyVoice-300M-SFT` `CosyVoice-300M-Instruct` model and `CosyVoice-ttsfrd` resource.
``` python
# SDK模型下载
from modelscope import snapshot_download
snapshot_download('FunAudioLLM/Fun-CosyVoice3-0.5B', local_dir='pretrained_models/Fun-CosyVoice3-0.5B')
snapshot_download('iic/CosyVoice2-0.5B', local_dir='pretrained_models/CosyVoice2-0.5B')
snapshot_download('iic/CosyVoice-300M', local_dir='pretrained_models/CosyVoice-300M')
snapshot_download('iic/CosyVoice-300M-SFT', local_dir='pretrained_models/CosyVoice-300M-SFT')
snapshot_download('iic/CosyVoice-300M-Instruct', local_dir='pretrained_models/CosyVoice-300M-Instruct')
snapshot_download('iic/CosyVoice-ttsfrd', local_dir='pretrained_models/CosyVoice-ttsfrd')
```
Optionally, you can unzip `ttsfrd` resource and install `ttsfrd` package for better text normalization performance.
Notice that this step is not necessary. If you do not install `ttsfrd` package, we will use wetext by default.
``` sh
cd pretrained_models/CosyVoice-ttsfrd/
unzip resource.zip -d .
pip install ttsfrd_dependency-0.1-py3-none-any.whl
pip install ttsfrd-0.4.2-cp310-cp310-linux_x86_64.whl
```
### Basic Usage
We strongly recommend using `CosyVoice3-0.5B` for better performance.
Follow the code in `example.py` for detailed usage of each model.
```sh
python example.py
```
#### CosyVoice2 vllm Usage
If you want to use vllm for inference, please install `vllm==v0.9.0`. Older vllm version do not support CosyVoice2 inference.
Notice that `vllm==v0.9.0` has a lot of specific requirements, for example `torch==2.7.0`. You can create a new env to in case your hardward do not support vllm and old env is corrupted.
``` sh
conda create -n cosyvoice_vllm --clone cosyvoice
conda activate cosyvoice_vllm
pip install vllm==v0.9.0 transformers==4.51.3 -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host=mirrors.aliyun.com
python vllm_example.py
```
#### Start web demo
You can use our web demo page to get familiar with CosyVoice quickly.
Please see the demo website for details.
``` python
# change iic/CosyVoice-300M-SFT for sft inference, or iic/CosyVoice-300M-Instruct for instruct inference
python3 webui.py --port 50000 --model_dir pretrained_models/CosyVoice-300M
```
#### Advanced Usage
For advanced users, we have provided training and inference scripts in `examples/libritts/cosyvoice/run.sh`.
#### Build for deployment
Optionally, if you want service deployment,
You can run the following steps.
``` sh
cd runtime/python
docker build -t cosyvoice:v1.0 .
# change iic/CosyVoice-300M to iic/CosyVoice-300M-Instruct if you want to use instruct inference
# for grpc usage
docker run -d --runtime=nvidia -p 50000:50000 cosyvoice:v1.0 /bin/bash -c "cd /opt/CosyVoice/CosyVoice/runtime/python/grpc && python3 server.py --port 50000 --max_conc 4 --model_dir iic/CosyVoice-300M && sleep infinity"
cd grpc && python3 client.py --port 50000 --mode
# for fastapi usage
docker run -d --runtime=nvidia -p 50000:50000 cosyvoice:v1.0 /bin/bash -c "cd /opt/CosyVoice/CosyVoice/runtime/python/fastapi && python3 server.py --port 50000 --model_dir iic/CosyVoice-300M && sleep infinity"
cd fastapi && python3 client.py --port 50000 --mode
```
#### Using Nvidia TensorRT-LLM for deployment
Using TensorRT-LLM to accelerate cosyvoice2 llm could give 4x acceleration comparing with huggingface transformers implementation.
To quick start:
``` sh
cd runtime/triton_trtllm
docker compose up -d
```
For more details, you could check [here](https://github.com/FunAudioLLM/CosyVoice/tree/main/runtime/triton_trtllm)
## Discussion & Communication
You can directly discuss on [Github Issues](https://github.com/FunAudioLLM/CosyVoice/issues).
You can also scan the QR code to join our official Dingding chat group.
## Acknowledge
1. We borrowed a lot of code from [FunASR](https://github.com/modelscope/FunASR).
2. We borrowed a lot of code from [FunCodec](https://github.com/modelscope/FunCodec).
3. We borrowed a lot of code from [Matcha-TTS](https://github.com/shivammehta25/Matcha-TTS).
4. We borrowed a lot of code from [AcademiCodec](https://github.com/yangdongchao/AcademiCodec).
5. We borrowed a lot of code from [WeNet](https://github.com/wenet-e2e/wenet).
## Citations
``` bibtex
@article{du2024cosyvoice,
title={Cosyvoice: A scalable multilingual zero-shot text-to-speech synthesizer based on supervised semantic tokens},
author={Du, Zhihao and Chen, Qian and Zhang, Shiliang and Hu, Kai and Lu, Heng and Yang, Yexin and Hu, Hangrui and Zheng, Siqi and Gu, Yue and Ma, Ziyang and others},
journal={arXiv preprint arXiv:2407.05407},
year={2024}
}
@article{du2024cosyvoice,
title={Cosyvoice 2: Scalable streaming speech synthesis with large language models},
author={Du, Zhihao and Wang, Yuxuan and Chen, Qian and Shi, Xian and Lv, Xiang and Zhao, Tianyu and Gao, Zhifu and Yang, Yexin and Gao, Changfeng and Wang, Hui and others},
journal={arXiv preprint arXiv:2412.10117},
year={2024}
}
@article{du2025cosyvoice,
title={CosyVoice 3: Towards In-the-wild Speech Generation via Scaling-up and Post-training},
author={Du, Zhihao and Gao, Changfeng and Wang, Yuxuan and Yu, Fan and Zhao, Tianyu and Wang, Hao and Lv, Xiang and Wang, Hui and Shi, Xian and An, Keyu and others},
journal={arXiv preprint arXiv:2505.17589},
year={2025}
}
@inproceedings{lyu2025build,
title={Build LLM-Based Zero-Shot Streaming TTS System with Cosyvoice},
author={Lyu, Xiang and Wang, Yuxuan and Zhao, Tianyu and Wang, Hao and Liu, Huadai and Du, Zhihao},
booktitle={ICASSP 2025-2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
pages={1--2},
year={2025},
organization={IEEE}
}
```
## Disclaimer
The content provided above is for academic purposes only and is intended to demonstrate technical capabilities. Some examples are sourced from the internet. If any content infringes on your rights, please contact us to request its removal.