ComfyUI-ChatTTS

ComfyUI-ChatTTS
★ 4

语音合成文本转语音ComfyUI集成批量处理
在ComfyUI中集成ChatTTS，实现高质量、可控的文本到语音生成，支持参数调节与批量处理，直接输出音频供ComfyUI使用。
💡 在ComfyUI流程中为多条文本生成可控语音文件。
🍴 2 Forks💻 Python🔄 2025-05-12
📦
网盘下载
复制链接后前往夸克网盘下载
https://pan.quark.cn/s/e58c8376a81b
📦 requirements.txt
numpy<2.0.0
numba
torch>=2.1.0
torchaudio
tqdm
vector_quantize_pytorch
transformers>=4.41.1
vocos
IPython
pybase16384
pynini==2.1.5;
sys_platform
==
'linux'
WeTextProcessing;
sys_platform
==
'linux'
nemo_text_processing;
sys_platform
==
'linux'
av
pydub
Basic TTS Workflow
📄 README

ComfyUI-ChatTTS

A ComfyUI integration for ChatTTS, enabling high-quality, controllable text-to-speech generation directly in your ComfyUI workflows.

Example Workflows

Basic Text-to-Speech

This simple workflow demonstrates basic text-to-speech conversion:

  • Load the ChatTTS model
  • Sample a random speaker voice
  • Convert text to speech
  • Preview the audio output

    • Features

  • High-Quality Voice Synthesis – Generate natural-sounding speech from text input
  • Voice Control – Sample random speakers or customize voice characteristics
  • Parameter Adjustment – Fine-tune temperature, top-P, top-K and other generation parameters
  • Batch Processing – Support for batch text processing through split_batch option
  • Seamless Integration – Works directly with ComfyUI’s audio nodes

    • Installation

    Prerequisites

  • A working installation of ComfyUI
  • Python 3.8+ with PyTorch installed

    • Using ComfyUI Manager (Recommended)

  • Install ComfyUI Manager
  • Search for “ChatTTS” and install

    • Manual Installation

  • Navigate to your ComfyUI’s custom_nodes directory
  • Clone this repository:

    • “`

    git clone https://github.com/neverbiasu/ComfyUI-ChatTTS

    “`

  • Install the requirements:

    • “`

    cd ComfyUI-ChatTTS

    pip install -r requirements.txt

    “`

    Model Setup

    ChatTTS models will be automatically downloaded when first used, or you can manually place them in:

    ComfyUI/models/chattts/

    The first time you run the ChatTTSLoader node, it will:

  • Check for existing models in the models/chattts directory
  • If none are found, download models from the official repository
  • Load the model for use in your workflows

    • ChatTTS Control Tags

    ChatTTS supports various special tags that can be inserted into your text to control the speech generation. These tags allow you to customize the speech output without changing the model parameters.

    | Tag | Range | Description |

    | ———— | —– | ——————————————————- |

    | [speed_n] | 1-9 | Controls speech speed (higher numbers = faster) |

    | [oral_n] | 0-9 | Controls oral expressiveness style |

    | [laugh_n] | 0-2 | Controls laughter intensity |

    | [break_n] | 0-7 | Controls pause duration (higher numbers = longer pause) |

    | [uv_break] | – | Inserts a brief pause/break at the word level |

    | [lbreak] | – | Inserts a longer pause/break (similar to line break) |

    | [laugh] | – | Inserts laughter at the specified position |

    Acknowledgements

  • ChatTTS for the core text-to-speech technology
  • ComfyUI for the wonderful UI framework

    • License

    This project is licensed under the MIT License – see the LICENSE file for details.