mcp-tavily

mcp-tavily

54

Tavily MCP Server is a Model Context Protocol server designed to enable AI-powered web search capabilities. It provides tools for comprehensive web searches and direct AI-generated answers, utilizing Tavily's search API. Key features include the ability to filter searches by domain and gather recent news articles.

Tavily MCP Server

A Model Context Protocol server that offers AI-powered web search capabilities through Tavily's search API, enabling sophisticated web searches and AI-generated answers to recent news articles.

Features

  • Tavily Web Search: Comprehensive web searches with AI content extraction.
  • Tavily Answer Search: Web searches generating direct AI-generated answers.
  • Tavily News Search: Recent news articles search with publication dates.

Prerequisites

  • Python 3.11 or later
  • Tavily API key
  • uv package manager (recommended)

Installation

Options for installation include using pip or uv, or from source.

Configuration

API key setup through .env files, environment variables, or command-line arguments.

Usage Examples

Search examples include basic web searches, domain-filtered reports, answer search mode, and news searches.

Testing

Guidelines for running a comprehensive test suite.

Docker

Guidelines for building and running a Docker container.

Debugging

Use the MCP inspector for debugging.

Contributing

Guidelines for contributing to the project.

License

Licensed under the MIT License.