1 required secret

Tavily Search

Real-time web research, data extraction, and domain crawling with AI ranking

Remote Updated 3 hours ago
Retrieval & SearchSaaS & API IntegrationsDeveloper Tools

About

Tavily Crawl Beta

GitHub Repo stars npm

MCP demo

A Model Context Protocol (MCP) server that provides comprehensive web research capabilities through fast, reliable APIs. It enables AI assistants to search the web in real-time, extract structured data from web pages, map website structures, and crawl entire domains with intelligent content extraction and ranking for AI applications.

Features

  • Real-time Web Search: Optimized search with intelligent content extraction and relevancy ranking for contextual information
  • Structured Data Extraction: Transform web pages into LLM-ready data with automatic content cleaning and format conversion
  • Website Mapping: Create structured maps of website architecture to understand organization and discover pages
  • Domain Crawling: Systematically explore websites with smart crawling strategies that respect robots.txt and handle dynamic content
  • Multiple Output Formats: Support for summaries, Markdown, or cleaned text extraction
  • Advanced Filtering: Control search scope with domain inclusion/exclusion, time ranges, and content categories

What you’ll need to connect

Required:

Configuration

Request headers

HeaderDescriptionRequired
AuthorizationTavily API KeyYes

Remote runtime

Fixed U R L
https://mcp.tavily.com/mcp/