Stats
Actions
Tags
From claude-soma
Local MCP server for speech-to-text transcription using Whisper models. Converts audio input to text via a Python backend.
Copy this JSON into your .mcp.json to enable this server
Add to your .mcp.json:
{
"mcpServers": {
"voice-stt": {
"command": "/opt/claude-soma/.venv/bin/python",
"args": [
"-m",
"claude_soma.mcp_servers.voice_stt.server"
],
"env": {
"HERMES_WHISPER_BIN": "/opt/whisper.cpp/build/bin/whisper-cli",
"HERMES_WHISPER_MODEL": "/opt/whisper.cpp/models/ggml-base.en.bin"
}
}
}
}Server configuration and connection parameters
/opt/claude-soma/.venv/bin/pythonCommand-line arguments passed to the server process
Environment variables set when the server starts
HERMES_WHISPER_BIN=/opt/whisper.cpp/build/bin/whisper-cliHERMES_WHISPER_MODEL=/opt/whisper.cpp/models/ggml-base.en.binnpx claudepluginhub techfreakworm/claude-soma