Content
# speech-mcp
Unified STT/TTS MCP-Server — speech-to-text via WhisperX, text-to-speech via ComfyUI/Qwen3, plus optional HTTP wrapper.
## Tools
| Tool | Description |
|---|---|
| `speech__record` | Record audio from USB microphone (N seconds) |
| `speech__transcribe` | Transcribe audio file via WhisperX |
| `speech__tts` | Text-to-speech via TTS endpoint |
## Setup
```bash
cd speech-mcp
python3 -m venv .venv
.venv/bin/pip install -r requirements.txt
cp .env.example .env # adjust WHISPERX_URL / TTS_URL / MIC_DEVICE
```
## MCP registration
```json
{
"mcpServers": {
"speech-mcp": {
"command": ".venv/bin/python",
"args": ["server.py"]
}
}
}
```
## HTTP mode (optional)
```bash
.venv/bin/python http_server.py
# Serves on http://localhost:8902
```
## Configuration
All config via environment variables (see `.env.example`):
| Variable | Default | Description |
|---|---|---|
| `WHISPERX_URL` | `localhost:48001` | WhisperX service endpoint |
| `TTS_URL` | `localhost:8188` | TTS service endpoint |
| `MIC_DEVICE` | `hw:1,0` | ALSA microphone device |
| `RECORD_DIR` | `/tmp/speech-mcp-recordings` | Recording scratch directory |
## License
MIT — see `LICENSE`.
