Tavily Search

v1.0.0Search & Data Extractionstable

An MCP server implementation that integrates the Tavily Search API, providing optimized search capabilities for LLMs.

tavily-searchmcpai-integration
Share:
47
Stars
0
Downloads
0
Weekly
0/5

What is Tavily Search?

Tavily Search is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to mcp server implementation that integrates the tavily search api, providing optimized search capabilities for llms.

An MCP server implementation that integrates the Tavily Search API, providing optimized search capabilities for LLMs.

This server falls under the Search & Data Extraction category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • An MCP server implementation that integrates the Tavily Sear

Use Cases

Perform optimized web searches for LLMs
Get high-quality search results integrated with AI agents
Tomatio13

Maintainer

LicenseMIT License
Languagepython
Versionv1.0.0
UpdatedMay 4, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

NPM

npx -y tavily-search

Manual Installation

npx -y tavily-search

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Tavily Search

The Tavily Search MCP server integrates the Tavily Search API into MCP-compatible AI clients, providing web search capabilities optimized for LLMs. Unlike general search engines, Tavily returns structured, AI-friendly results including an AI-generated summary answer alongside ranked source URLs — making it particularly useful for research tasks where Claude needs current web information. The server exposes a single 'search' tool with support for both basic and advanced search depth modes.

Prerequisites

  • A Tavily API key — sign up at https://tavily.com to get one
  • Python 3.10 or later with uv installed, or Docker for containerized deployment
  • The mcp-server-tavily source code cloned from GitHub
  • An MCP-compatible client such as Claude Desktop
1

Obtain a Tavily API key

Sign up at https://tavily.com and retrieve your API key from the dashboard. Tavily offers a free tier suitable for personal use.

2

Clone the server repository

Clone the mcp-server-tavily repository to your local machine.

git clone https://github.com/Tomatio13/mcp-server-tavily.git
cd mcp-server-tavily
3

Install uv (if not already installed)

uv is used to run the Python server without manually creating a virtual environment.

curl -LsSf https://astral.sh/uv/install.sh | sh
4

Test the server locally

Run the server once to confirm it starts correctly with your API key.

TAVILY_API_KEY="your_tavily_api_key" uv --directory /path/to/mcp-server-tavily run tavily-search
5

Add to your MCP client configuration

Edit your claude_desktop_config.json to register the Tavily search server. Set the directory path to where you cloned the repo.

{
  "mcpServers": {
    "tavily-search": {
      "command": "uv",
      "args": ["--directory", "/path/to/mcp-server-tavily", "run", "tavily-search"],
      "env": {
        "TAVILY_API_KEY": "your_tavily_api_key",
        "PYTHONIOENCODING": "utf-8"
      }
    }
  }
}
6

Restart Claude Desktop

Fully quit and reopen Claude Desktop to load the new server configuration. You can verify it's active by asking Claude to search for something.

Tavily Search Examples

Client configuration

Add this block to your claude_desktop_config.json, replacing the directory path with the actual location of your cloned mcp-server-tavily folder.

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

Prompts to try

After connecting the Tavily Search server, use these prompts to perform web searches through Claude.

- "Search for the latest news on OpenAI's newest model releases"
- "Find recent research papers on transformer architecture improvements"
- "What are today's top headlines in AI?"
- "Search for the current price of NVIDIA stock"
- "Find documentation for the Tailwind CSS v4 release"

Troubleshooting Tavily Search

Authentication error — TAVILY_API_KEY is invalid or missing

Double-check your API key from the Tavily dashboard at https://app.tavily.com. Ensure it is set in the env block of your MCP config, not hardcoded in the args array.

Garbled or non-UTF-8 characters in search results

Add PYTHONIOENCODING=utf-8 to the env section of your config. This is especially important on Windows systems or when searching for non-ASCII content.

uv command not found when Claude Desktop tries to start the server

Install uv globally by running 'curl -LsSf https://astral.sh/uv/install.sh | sh' and reload your shell. Make sure uv is on the system PATH that Claude Desktop uses (on macOS, add it to /etc/paths or your shell profile).

Frequently Asked Questions about Tavily Search

What is Tavily Search?

Tavily Search is a Model Context Protocol (MCP) server that mcp server implementation that integrates the tavily search api, providing optimized search capabilities for llms. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Tavily Search?

Install via npm with the command: npx -y tavily-search. Then add the server configuration to your AI client's JSON config file (e.g., claude_desktop_config.json or .cursor/mcp.json).

Which AI clients work with Tavily Search?

Tavily Search works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Tavily Search free to use?

Yes, Tavily Search is open source and available under the MIT License license. You can use it freely in both personal and commercial projects.

Browse More Search & Data Extraction MCP Servers

Explore all search & data extraction servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "tavily-search": { "command": "npx", "args": ["-y", "tavily-search"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Tavily Search?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides