home / mcp / tavily mcp server
Production ready MCP server with real-time search, extract, map & crawl.
Configuration
View docs{
"mcpServers": {
"tavily-ai-tavily-mcp": {
"url": "https://mcp.tavily.com/mcp/?tavilyApiKey=<your-api-key>"
}
}
}The Tavily MCP Server empowers you to search, extract, map, and crawl web data with real-time capabilities. It provides dedicated tools for on-demand data gathering and structured website mapping, enabling you to build powerful data pipelines and integrations with ease.
You can connect to Tavily MCP in two ways: use the remote Tavily MCP service or run the MCP server locally. The remote service lets you access Tavily’s search, extract, map, and crawl capabilities without any local installation. Running locally gives you full control on your own machine.
To connect to Tavily’s remote MCP server, you provide the remote URL and your Tavily API key. You can also authenticate using an Authorization header if your client supports it. When you use the remote service, you can optionally set default request parameters that apply to all requests.
You can run Tavily MCP locally using NPX. This starts a local MCP instance on your machine, giving you direct access to Tavily’s tools from your environment.
- Real-time web search capabilities through the Tavily search tool. - Intelligent data extraction from web pages via Tavily extract. - Web mapping that creates a structured map of websites. - A web crawler that systematically explores websites to gather data.
If you want to standardize behavior across requests, you can set default parameters for the Tavily search tool. For remote connections, your API key should be kept secure, and you can choose to pass it via the URL or an Authorization header depending on your client’s capabilities.
Performs real-time web searches using Tavily’s capabilities to fetch relevant results based on your queries.
Intelligently extracts data from web pages, transforming raw content into structured data.
Creates a structured map of websites to visualize relationships and hierarchies between pages.
Systematically crawls websites to discover and index pages for data collection.
Tool for performing real-time searches across the web as part of the MCP suite.
Tool for extracting structured information from web pages as part of data gathering.