MCP Voice Assistant

MCP Voice Assistant is an MCP server and AI agent that connects Model Context Protocol tools to speech interfaces. It uses OpenAI Whisper for speech-to-text input and ElevenLabs alongside a pyttsx3 fallback for text-to-speech output. Built for developers, operators, and power users who prefer voice-driven workflows, this assistant allows users to control external systems without manual typing. By bridging audio processing with tool-calling language models such as OpenAI, Anthropic Claude, or Groq LLama, the assistant routes verbal instructions directly to active MCP integrations. Through an extensible configuration file, it coordinates diverse servers including Playwright for automated browser tasks, Linear for issue tracking, local filesystems, GitHub, and database connectors. The assistant maintains conversational context across prompts, permitting dynamic multi-step tasks to be completed entirely via spoken dialogue.

Category: Communication & Messaging

Tags: assistant, audio, speech-to-text, text to speech, voice

Visit MCP Voice Assistant

How to install and configure MCP Voice Assistant

1. Install system audio dependencies (such as brew install portaudio on macOS or sudo apt-get install portaudio19-dev on Ubuntu). 2. Clone the repository and navigate into the folder: git clone https://github.com/yourusername/mcp-voice-assistant.git && cd mcp-voice-assistant 3. Create and activate a virtual environment using uv: uv venv && source .venv/bin/activate 4. Install the package: uv pip install -e . 5. Set up required keys in a .env file, such as OPENAI_API_KEY and optional variables like ELEVENLABS_API_KEY. 6. Configure your tool servers in mcp_servers.json. 7. Run the agent using python voice_assistant/agent.py or through custom CLI arguments.

What you can do with MCP Voice Assistant

• Automate web workflows by issuing spoken instructions to trigger Playwright browser scripts and web scraping. - Manage project boards hands-free by voice-dictating tickets, task updates, and priority shifts into Linear. - Query local file paths or inspect project directory structures without switching away from active terminal displays. - Run database inspections or execute standard queries using voice commands coupled to integrated SQL MCP servers. - Interact with coding repositories hands-free to review issues or inspect pull requests via GitHub integrations.

Key facts

• https://github.com/yourusername/mcp-voice-assistant
• Communication & Messaging, Design, Media & Creative
• assistant, audio, speech-to-text, text to speech, voice

Part of MCP Servers

How do I install MCP Voice Assistant?

Install PortAudio on your operating system, clone the repository, and set up a virtual environment using uv. Install the package via uv pip install, provide your API keys in a .env file, configure mcp_servers.json, and launch the assistant script.

What LLM providers work with MCP Voice Assistant?

The project supports any language model that provides tool-calling capabilities through LangChain. Tested providers include OpenAI with GPT-4 or GPT-3.5, Anthropic Claude models, and Groq with LLama 3.

What voice technologies does MCP Voice Assistant use?

The assistant uses OpenAI Whisper for real-time speech-to-text input processing. Spoken responses are generated using ElevenLabs text-to-speech, with an automatic fallback to pyttsx3 if ElevenLabs is unavailable or fails.

Which MCP servers can MCP Voice Assistant connect to?

It comes preconfigured for Playwright and Linear. You can add additional servers by editing mcp_servers.json, including filesystem, GitHub, GitLab, Google Drive, PostgreSQL, SQLite, Slack, Brave Search, and Fetch servers.

Is MCP Voice Assistant open source?

Yes, MCP Voice Assistant is open-source software licensed under the MIT License. You can view the repository, contribute improvements, or fork the codebase directly on GitHub.