Gemini 2.5 Pro MCP Servers: Automate ANYTHING!

Key Concepts

• MCP (Multi-Chain Protocol) Servers: Enable AI agents to access and utilize functionalities from various applications and services.
• Gemini 2.5 Pro: Google's AI model used for coding and building MCP servers.
• Visual Studio Code: A free code editor used for setting up and managing MCP servers.
• Client (VS Code Extension): A free extension within Visual Studio Code that simplifies MCP server setup.
• API Key: A code used to authenticate and authorize access to services like Gemini 2.5 Pro, Open Router, and Plexi.
• Open Router: A platform for accessing various AI models through a single API.
• AI Agents: Software programs that use AI to perform tasks autonomously.

Setting up MCP Servers with Gemini 2.5 Pro and Client

1. Introduction to MCP Servers and Gemini 2.5 Pro

• Gemini 2.5 Pro allows building MCP servers for coding apps, games, and tools.
• MCP servers are the future of AI agents, enabling them to connect to the internet and access multiple functionalities efficiently.
• Traditional AI agent development requires setting up individual tools and apps for each feature, which is time-consuming.
• MCP servers allow an AI agent to link to a single MCP (e.g., Gmail MCP) and access all its functionalities directly.

2. Setting up the Environment

• Visual Studio Code: Download and install Visual Studio Code (free).
• API Key: Obtain an API key from AIS.google.com (Gemini 2.5 Pro is free).
• Client Extension: Install the "Client" extension in Visual Studio Code (free).

3. Installing and Configuring Client

• Open Visual Studio Code and go to the extensions tab.
• Search for "Client" and install the extension.
• Click on the Client icon in the activity bar.
• Go to "Settings" within the Client extension.
• Navigate to the MCP icon to access the marketplace for setting up MCP servers.

4. Installing MCP Servers from the Client Marketplace

• The Client marketplace provides a simplified way to install MCP servers compared to methods like Cursor.
• Search for the desired MCP server (e.g., Plexi Research) and click "Install."
• Client will use an API request to set up the MCP server.
• Provide the necessary API key (e.g., Plexi API key) when prompted.

5. Troubleshooting and Best Practices

• If you encounter issues, ensure you are using a new folder in Visual Studio Code.
• Using Gemini Pro Experimental directly can simplify the setup process.
• Be patient, as MCP servers are a relatively new technology and may be temperamental.
• Some MCP servers may not work perfectly (e.g., Browser use).

6. Testing the Installed MCP Server

• After installation, test the MCP server by sending a query (e.g., "What happened in AI news today?" for Plexi).
• The AI agent will use the MCP server to retrieve and format the information.

7. Exploring Available MCP Servers

• Check the MCP server's GitHub repository to understand its functionalities and available API keys.
• Examples of available MCP servers include Google Drive, Google Maps, Slack, SendGrid, Wolfram Alpha, and Spotify.

8. Example: Plexi Research MCP Server

• Plexi Research MCP server allows the AI agent to search for information using Plexi's AI-powered search engine.
• It provides AI-generated results, not just standard Google search results.

9. Example: Browser Use MCP Server

• The Browser Use MCP server aims to enable AI-driven browser automation.
• The presenter found it unreliable in the video.

10. Spotify MCP Server

• Requires Spotify API credentials obtained from the Spotify developer dashboard.

Conclusion

The video demonstrates how to build MCP servers using Google Gemini 2.5 Pro and the Client extension in Visual Studio Code. It highlights the benefits of MCP servers for AI agents, simplifying access to various functionalities. The presenter provides a step-by-step guide, troubleshooting tips, and examples of available MCP servers. The key takeaway is that using the Client extension with Gemini Pro Experimental in a new Visual Studio Code folder offers the simplest and most efficient way to set up MCP servers. While some MCP servers may be unreliable, others like Plexi Research can significantly enhance the capabilities of AI agents.