Kokoro Tts MCP

🚀 Kokoro Text-to-Speech (TTS) MCP Server

The Kokoro Text-to-Speech MCP server is used to generate .mp3 files and provides an option to upload them to S3.

Usage: https://huggingface.co/spaces/hexgrad/Kokoro-TTS

🚀 Quick Start

✨ Features

• Generate .mp3 files from text.
• Provide an option to upload the generated .mp3 files to S3.

📦 Installation

Clone the Repository

Clone the repository to your local machine.

Download Model Files

Download kokoro-v1.0.onnx and voices-v1.0.bin from Kokoro Onnx weights and store them in the same repository.

Configure MCP

Add the following content to your MCP configuration and update it with your own values.

  "kokoro-tts-mcp": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/toyourlocal/kokoro-tts-mcp",
        "run",
        "mcp-tts.py"
      ],
      "env": {
        "TTS_VOICE": "af_heart",
        "TTS_SPEED": "1.0",
        "TTS_LANGUAGE": "en-us",
        "AWS_ACCESS_KEY_ID": "",
        "AWS_SECRET_ACCESS_KEY": "",
        "AWS_REGION": "us-east-1",
        "AWS_S3_FOLDER": "mp3",
        "S3_ENABLED": "true",
        "MP3_FOLDER": "/path/to/mp3"
      } 
    }

Install ffmpeg

This is required to convert .wav files to .mp3 files.

For Mac:

brew install ffmpeg

Configure Environment Variables

To run locally, add these contents to your .env file. Refer to env.example, copy it to .env, and modify it according to your own values.

💻 Usage Examples

Run the Server Locally

The recommended method is to use UV:

uv run mcp-tts.py

Text-to-Speech Client

To use the text-to-speech function, run the following command in the terminal:

uv run mcp-client.py --tts text-to-speech <your-text>

Replace <your-text> with the actual text content.

Example: Configure and Run the MCP Server and Client

1. Start the MCP server:

uv run mcp-tts.py

2. In another terminal window, start the TTS client:

uv run mcp-client.py --tts text-to-speech "你好，Kokoro！"

This will use the Kokoro model to convert Chinese text to speech and generate an .mp3 file.

📚 Documentation

Supported Environment Variables

🔧 Technical Details

Customization and Extension

To customize the behavior of the MCP server, edit the configuration parameters in mcp-tts.py. For example, you can adjust the audio output format, sampling rate, or bit depth.

To add more features or integrate other services, extend the code of the MCP server by adding new handlers and services.

Text-to-Speech Example

import uvicorn
from fastapi import FastAPI
from fastapi.responses import FileResponse
from pydub import AudioSegment
import os

app = FastAPI()

@app.get("/text-to-speech/{text}")
async def text_to_speech(text: str):
    # Here it is assumed that you have the logic to generate speech, such as using the API of iFlytek or Alibaba Cloud
    # The following is an example, and the actual implementation needs to be adjusted according to the specific situation
    audio = AudioSegment.from_text(text, language="zh-CN")
    file_path = os.path.join("audio_files", f"output_{len(audio)}_bytes.mp3")
    audio.export(file_path, format="mp3")
    return FileResponse(file_path)

if __name__ == "__main__":
    uvicorn.run(app, host="0.0.0.0", port=8000)

Other Resources

• FastAPI Official Documentation
• Pydub Documentation
• iFlytek Speech Synthesis API
• Alibaba Cloud Speech Synthesis Service