Local Tts

A lightweight, local Model Context Protocol (MCP) server that exposes a text-to-speech tool. An AI assistant (such as Claude Code) can send its written task summaries to the tool, which reads them aloud instantly using your computer's native, offline speech engine.

Because the speech engine (pyttsx3) runs entirely on your machine, there is near-zero latency and no network calls.

Features

• speak_text(text, rate=175, volume=1.0) — read text aloud through the native OS voice. rate (words per minute) and volume (0.0–1.0) are optional overrides.
• list_voices() — enumerate the voices installed on the local system.

Requirements

• Python 3.10+
• A working native speech engine:
  • macOS: built in (NSSpeechSynthesizer) — no setup needed.
  • Windows: built in (SAPI5) — no setup needed.
  • Linux: install espeak/espeak-ng and ALSA, e.g. sudo apt-get install espeak-ng libespeak1.

Setup

# 1. Clone and enter the project
cd tts-mcp

# 2. Create and activate a virtual environment
python -m venv venv

# macOS/Linux:
source venv/bin/activate
# Windows (Command Prompt):
venv\Scripts\activate.bat
# Windows (PowerShell):
.\venv\Scripts\Activate.ps1

# 3. Install dependencies
pip install -r requirements.txt

Local testing

Verify the server starts without errors:

python server.py

It runs over stdio and stays active. Press Ctrl+C to stop it.

To exercise the tools interactively, use the MCP Inspector:

mcp dev server.py

Integrate with Claude Code / other MCP clients

Add the server to your MCP configuration (e.g. .mcp.json in your project or your client's global config). See mcp.json.example:

{
  "mcpServers": {
    "local-tts": {
      "command": "/absolute/path/to/tts-mcp/venv/bin/python",
      "args": ["/absolute/path/to/tts-mcp/server.py"]
    }
  }
}

Replace /absolute/path/to/ with your actual local paths. On Windows the command is typically ...\venv\Scripts\python.exe.

You can also register it with the Claude Code CLI:

claude mcp add local-tts /absolute/path/to/tts-mcp/venv/bin/python /absolute/path/to/tts-mcp/server.py

Prompting strategy

Tell your assistant that when a task is finished it should structure its response into concise bullet points and call speak_text with that summary — for example: "When you say Task complete, summarize the work as short bullet points and read it aloud with the speak_text tool."

License

MIT