Skip to content

Tavily MCP

Real-time web search, page extraction, site mapping and crawling for research.

Officialby TavilyHosted or on your computerChecked October 18, 2026

Tavily's official server gives Claude a search API built for AI agents, together with tools that extract page content, map a website's structure and crawl it. You can connect the hosted endpoint by signing in or by passing an API key, or run the open-source server locally with npx.

Who it is for

  • Researchers and analysts
  • Content teams checking facts
  • Developers building research agents

Prompts to try

Once it is connected, ask in plain words. Claude decides when to use the server and asks before it changes anything important.

Prompt 1
Search for news about the latest interest rate decision and summarize the market reaction with sources.
Prompt 2
Map the structure of competitor.com and tell me which product categories they have.
Prompt 3
Extract the text of these three articles and compare their conclusions.

How to connect it

In Claude Code

  1. claude mcp add --transport http tavily https://mcp.tavily.com/mcp

In the Claude desktop app

Customize > Connectors > Discover: search Tavily, then select Connect to Claude

Hosted address: https://mcp.tavily.com/mcp

Good to know

Run /mcp in Claude Code to finish sign-in, or append ?tavilyApiKey=YOUR_API_KEY to the URL instead.

Unlock the setup steps

Free. Leave your email once to open the setup steps for all 150 servers, and get the best new connectors each month.

No spam. One click to unsubscribe. We use your email to send you the course and AI tips.

How we checked it

Checked the GitHub repo (MIT, last commit 2026-10-15), Tavily's MCP docs, the tavily-mcp npm package and the Claude connectors listing; the hosted endpoint answers with an OAuth challenge.

Source: github.com/tavily-ai/tavily-mcp