Tavily Search MCP Server

Overview

Tavily Search is an MCP server that connects to the Tavily API to perform web searches based on your queries. It returns comprehensive search results including AI-generated summaries, source URLs, and titles of the search results.

Prerequisites

Before installing Tavily Search, you need to:

1. Obtain a Tavily API key from Tavily's website
2. Have Python installed on your system

Installation Options

Option 1: Installing via Smithery (Recommended)

The easiest way to install Tavily Search is through Smithery:

npx -y @smithery/cli install tavily-search --client claude

Option 2: Manual Installation

1. Clone the repository:

git clone https://github.com/Tomatio13/mcp-server-tavily.git

2. Navigate to the Claude Desktop configuration file:

   • On MacOS: ~/Library/Application\ Support/Claude/claude_desktop_config.json
   • On Windows: C:\Users\[username]\AppData\Roaming\Claude\claude_desktop_config.json

3. Add the following configuration to the JSON file:

"mcpServers": {
  "tavily-search": {
    "command": "uv",
    "args": [
      "--directory",
      "PATH_TO_REPOSITORY/mcp-server-tavily",
      "run",
      "tavily-search"
    ],
    "env": {
      "TAVILY_API_KEY": "YOUR_TAVILY_API_KEY",
      "PYTHONIOENCODING": "utf-8"
    }
  }
}

4. Replace PATH_TO_REPOSITORY with the actual path where you cloned the repository and YOUR_TAVILY_API_KEY with your actual Tavily API key.

5. Restart Claude Desktop.

Option 3: Installation for Cursor

1. Create a shell script (e.g., tavily-search.sh) with the following content:

#!/bin/bash
TARGET_DIR=/path/to/mcp-server-tavily
cd "${TARGET_DIR}"
export TAVILY_API_KEY="your-api-key"
export PYTHONIOENCODING=utf-8
uv --directory $PWD run tavily-search

2. Make the script executable:

chmod +x tavily-search.sh

3. Configure Cursor's MCP Server settings:

   • Name: tavily-search
   • Type: command
   • Command: /path/to/your/tavily-search.sh

4. Save the settings and restart Cursor.

Option 4: Using Docker Compose

For operating systems where Claude Desktop isn't available:

1. Install Docker on your system
2. Clone the repository:

git clone https://github.com/Tomatio13/mcp-server-tavily.git

3. Run Docker compose:

docker compose up -d

4. Execute the client:

docker exec mcp_server uv --directory /usr/src/app/mcp-server-tavily/src run client.py

Usage

Once installed, you can use Tavily Search by asking your AI assistant to search for information. For example:

• "Please search for today's events in Tokyo"
• "Search for the latest news about climate change"
• "Find information about the history of the Eiffel Tower"

The search results will include:

• An AI-generated summary of the search results
• Source URLs and titles for reference

Logs

Logs are stored in the following locations:

• Windows: C:\Users\[username]\AppData\Roaming\Claude\logs\mcp-server-tavily-search
• MacOS: Check the Claude Desktop logs directory

Troubleshooting

If you encounter issues:

1. Verify your Tavily API key is correct
2. Check the logs for error messages
3. Ensure the paths in your configuration are correct
4. Restart the Claude Desktop or Cursor application