The Complete Guide to MCP: Giving Your AI Web Superpowers

With Model Context Protocol (MCP) servers, you can give even lightweight, locally-run models the ability to search the web, analyze articles, and access real-time data—all while maintaining complete privacy and without API fees.

Tools like Brave Search, Tavily, and DuckDuckGo offer generous free tiers, making it possible to create a private “SearchGPT” on your own machine.

Core Concepts

• Model Context Protocol (MCP): An open standard that allows AI models to connect with external tools and data sources. Instead of a rigid API call, MCP allows the model to determine which tool to use to fulfill a user’s need.
• Tool Calling (Function Calling): The mechanism that enables an AI model to recognize when it needs external information and invoke the appropriate function (or MCP tool) to get it. Without tool-calling support, a model is limited to its training data.

Here’s how to give these superpowers to your local model.

Technical Requirements and Setup

• Node.js: Must be installed on your computer.
• Python: Should be installed.
• Local AI Application: An application that supports MCP, such as LM Studio (version 0.3.17+), Claude Desktop, or Cursor IDE.
• Tool-Calling Model: A local model with tool-calling capabilities.

Good examples of local models that support tool-calling include GPT-oss, DeepSeek R1, Jan-v1-4b, and Llama-3.2 Instruct. In LM Studio, these models are often marked with a hammer icon.

Setting Up Search Engines

Configuration is managed through a single mcp.json file. In LM Studio, you can access this via the settings interface. Each MCP server entry requires a name, the command to run it, and any environment variables (like API keys).

DuckDuckGo (Easiest Setup)

1. Go to lmstudio.ai/danielsig/duckduckgo and click “Run in LM Studio.”
2. Go to lmstudio.ai/danielsig/visit-website and do the same.

Your model now has basic web search and page-reading capabilities.

Brave Search (Privacy-Focused)

Brave Search runs on an independent index and offers 2,000 free queries per month.

1. Sign up at brave.com/search/api to get your API key.
2. In LM Studio, go to Settings (wrench icon) > Program > Edit mcp.json.
3. Paste the following configuration, replacing the placeholder with your API key:

{
  "mcpServers": {
    "brave-search": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-brave-search"],
      "env": {
        "BRAVE_API_KEY": "your_brave_api_key_here"
      }
    }
  }
}

Tavily (Versatile Search)

Tavily provides 1,000 free credits per month and offers specialized search for news, code, and images.

1. Create an account at app.tavily.com and generate your API key.
2. Paste the following into your mcp.json, replacing the placeholder with your key:

{
  "mcpServers": {
    "tavily-remote": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://mcp.tavily.com/mcp/?tavilyApiKey=YOUR_API_KEY_HERE"]
    }
  }
}

Reading and Interacting with Websites

MCP Fetch (Reading Full Articles)

Search engines return snippets, but MCP Fetch retrieves the complete content of a webpage and converts it to AI-friendly markdown.

1. Install uvx, a Python package installer: pip install uvx.
2. Add this configuration to your mcp.json:

{
  "mcpServers": {
    "fetch": {
      "command": "uvx",
      "args": [
        "mcp-server-fetch"
      ]
    }
  }
}

You can now ask your model to summarize or analyze entire articles by providing a URL.

MCP Browser (Full Interaction)

Tools like MCP Browser or Playwright allow the model to interact with websites, including filling forms and navigating JavaScript-heavy pages.

The Complete Configuration

Here is a complete mcp.json file integrating all the services mentioned. Replace the API key placeholders, save the file, and restart your AI application.

{
  "mcpServers": {
    "fetch": {
      "command": "uvx",
      "args": [
        "mcp-server-fetch"
      ]
    },
    "brave-search": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-brave-search"
      ],
      "env": {
        "BRAVE_API_KEY": "YOUR_BRAVE_API_KEY_HERE"
      }
    },
    "browsermcp": {
      "command": "npx",
      "args": [
        "@browsermcp/mcp@latest"
      ]
    },
    "tavily-remote": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://mcp.tavily.com/mcp/?tavilyApiKey=YOUR_TAVILY_API_KEY_HERE"
      ]
    }
  }
}

This configuration provides robust, private, and free web access for your local models.