Doubao TTS

本项目旨在将火山引擎TTS官方API封装为标准MCP协议工具，支持在 Cursor、Claude 等平台通过自然语言参数调用，实现一键语音合成。适用于需要自定义音色、参数映射、自动化批量语音生成的开发者场景。

python

GitHub Stars

python

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "lxy2109-doubao-tts-mcp": {
      "command": "doubao-tts-mcp",
      "args": [],
      "env": {
        "PORT": "5001",
        "OUTPUT_DIR": "/path/to/output",
        "VOLC_APPID": "YOUR_APP_ID",
        "VOLC_TOKEN": "YOUR_TOKEN"
      }
    }
  }
}

You run a local MCP server that wraps the Volcengine TTS API for easy voice synthesis from any MCP-enabled platform. It lets you customize voice styles, adjust parameters, and generate audio files in bulk from natural language inputs, all through a simple command-line interface and a shared MCP configuration.

How to use

To use this MCP server, install it locally, start the service, and connect your MCP client (Cursor, Claude, or any MCP-compatible platform). The client will pass voice parameters and text to the local MCP server, which returns the path to the synthesized audio file. You can adjust voice type to map to official voice models, tweak speed, and set an output directory for generated files.

How to install

Prerequisites: You need Python 3.8 or newer.

Step 1. Install Python and verify version by running a version check.

pip install -r requirements.txt

Step 2. Install the local package so you can call it from the command line.

pip install -e .

Step 3. Configure environment variables. You can place these in a .env file or inside mcp.json under env. Required values are provided by your Volcengine TTS account.

VOLC_APPID=your_app_id
VOLC_TOKEN=your_token
PORT=5001
OUTPUT_DIR=/path/to/output

Step 4. Start the MCP service locally.

doubao-tts-mcp

Step 5. If you are using Cursor or Claude, provide a mcp.json configuration that points to the local command so the client can launch the server automatically.

{
  "mcpServers": {
    "doubao_tts_mcp": {
      "command": "doubao-tts-mcp",
      "args": [],
      "env": {
        "VOLC_APPID": "你的appid",
        "VOLC_TOKEN": "你的token",
        "PORT": "5001",
        "OUTPUT_DIR": "D:/doubao_tts_mcp/output"
      }
    }
  }
}

Additional notes

在 Cursor/Claude 等平台中，MCP 客户端会自动读取你的配置并以命令行形式启动本地 MCP 服务。你在 MCP 面板填写参数后即可一键生成音频，合成结果通常会返回音频文件的绝对路径，音频文件会保存在你指定的 OUTPUT_DIR 目录下。

关于参数的取值：你可以使用自然语言关键词来描述音色、情感和语速，例如输入“少女”或“萝莉”来自动映射到官方音色。比特率和采样率等参数在当前工具中暂不可用，请使用默认值以确保兼容性。

示例请求字段（用于理解如何在客户端传参）：文本、声音类型、语速、情感、输出文件名等。

{
  text: "测试声音参数",
  voice_type: "少女",
  speed_ratio: 0.8,
  emotion: "happy",
  output_filename: "少女音色测试"
}

Built by

VeilStrat

AI signals for GTM teams