Speaking Bots MCP Server

What is Speaking Bots MCP Server?

Deploy AI-powered speaking agents in video meetings using the MeetingBaas API. This open-source MCP server connects LLM-powered voice bots to Google Meet, Zoom, and Microsoft Teams, with customizable personas and real-time audio streaming.

How to use Speaking Bots MCP Server?

Clone the repository, install dependencies, set the required environment variables (BAAS_API_KEY, BAAS_URL), then run locally or deploy to a platform like Docker or fly.io. Use the API endpoints (e.g., POST /bots) via curl or any HTTP client to join meetings as a speaking bot.

Key features of Speaking Bots MCP Server

• Join meetings as an AI speaking bot
• Customizable personas and voices
• Real-time bidirectional audio streaming
• Minimalist, extensible API design
• Open source and easy to customize
• Integrates with MeetingBaas and Pipecat

Use cases of Speaking Bots MCP Server

• Automate onboarding in virtual meetings with the baas_onboarder persona
• Add creative role-play characters (e.g., pirate_queen) to team calls
• Deploy a quantum physicist persona for educational demonstrations
• Build custom AI assistants for recurring meeting workflows

FAQ from Speaking Bots MCP Server

What frameworks does Speaking Bots MCP Server use?

It uses Pipecat for building voice and multimodal conversational agents, and the MeetingBaas API for video meeting integration.

How do I set up the server?

Clone the repo, install dependencies, set BAAS_API_KEY and BAAS_URL in your environment, then run locally or deploy (e.g., Docker, fly.io, Vercel).

What meeting platforms are supported?

Google Meet, Zoom, and Microsoft Teams are explicitly supported.

How do I authenticate API requests?

Include your MeetingBaas API key in the x-meeting-baas-api-key header on each request.

What are the main API endpoints?

The key endpoints are POST /bots (join a meeting), DELETE /bots/{bot_id} (remove a bot), and POST /personas/generate-image (generate a persona image).