Tavily Search MCP Server

SearchTypeScript

Access real-time web search, data extraction, site mapping, and web crawling capabilities

Available Tools

tavily-search

Performs real-time web searches with customizable parameters like search type, max results, and domain filtering

querysearch_depthinclude_domainsexclude_domainsinclude_answersearch_typemax_results

tavily-extract

Extracts structured data from web pages with customizable extraction parameters

urlextraction_typeextraction_params

tavily-map

Creates a structured map of a website's content and structure

urlmax_pagesmax_depth

tavily-crawl

Systematically explores websites to gather information with customizable crawling parameters

urlmax_pagesmax_depth

Tavily Search provides AI assistants with powerful web search and data extraction capabilities through the Model Context Protocol (MCP). It enables real-time access to web information with sophisticated filtering options, intelligent data extraction from web pages, structured website mapping, and systematic web crawling functionality.

Overview

Tavily Search is an MCP server that connects AI assistants to the web, providing real-time search capabilities, data extraction, website mapping, and web crawling functionality. This integration allows AI models like Claude to access up-to-date information from the internet and extract structured data from websites.

Prerequisites

Before using Tavily Search, you'll need:

A Tavily API key (sign up for a free account if you don't have one)
Node.js v20 or higher
An MCP-compatible client like Claude Desktop, Cursor, or VS Code with MCP support

Installation Options

Quick Start with NPX

The simplest way to run Tavily Search is using NPX:

npx -y tavily-mcp@0.2.0

Using Smithery (for Claude Desktop)

For automatic installation via Smithery:

npx -y @smithery/cli install @tavily-ai/tavily-mcp --client claude

Client Configuration

VS Code Configuration

You can add Tavily Search to VS Code by adding this to your User Settings (JSON):

{
  "mcp": {
    "inputs": [
      {
        "type": "promptString",
        "id": "tavily_api_key",
        "description": "Tavily API Key",
        "password": true
      }
    ],
    "servers": {
      "tavily": {
        "command": "npx",
        "args": ["-y", "tavily-mcp@0.2.0"],
        "env": {
          "TAVILY_API_KEY": "${input:tavily_api_key}"
        }
      }
    }
  }
}

Alternatively, you can create a .vscode/mcp.json file in your workspace with the same configuration.

Cline Configuration

To use Tavily Search with Cline in VS Code:

Open Cline in VS Code
Navigate to the Marketplace tab
Find and install Tavily Search
Enter your Tavily API key when prompted

Claude Desktop Configuration

To configure Claude Desktop:

Open Claude Desktop
Go to Settings > Model Context Protocol
Add a new server with the following configuration:
- Name: Tavily
- Command: npx
- Arguments: -y tavily-mcp@0.2.0
- Environment Variables: TAVILY_API_KEY=your_api_key_here

Cursor Configuration

To configure Cursor:

Open Cursor
Go to Settings > AI > MCP Servers
Add a new server with:
- Name: Tavily
- Command: npx
- Arguments: -y tavily-mcp@0.2.0
- Environment Variables: TAVILY_API_KEY=your_api_key_here

Advanced Configuration

You can customize Tavily Search behavior using environment variables:

TAVILY_API_KEY: Your Tavily API key (required)
PORT: Server port (default: 8000)
HOST: Server host (default: localhost)

Troubleshooting

If you encounter issues:

Verify your Tavily API key is valid
Ensure Node.js v20+ is installed
Check for any error messages in the terminal where the MCP server is running
Make sure your client is properly configured to connect to the MCP server

For additional help, visit the Tavily documentation or check the GitHub repository for updates.