Skip to content

MCP setupWorkflow

Tavily MCP Server

Run Tavily's official MCP server through npx so your agent can search the web and extract page content with tavily-search and tavily-extract.

npx workspacetemplates add tavily-mcp --tool claude-code

Writes 1 file to your project root

.mcp.jsonmerge into your file14 lines · 208 BFile
{
  "mcpServers": {
    "tavily": {
      "command": "npx",
      "args": [
        "-y",
        "tavily-mcp@latest"
      ],
      "env": {
        "TAVILY_API_KEY": "${TAVILY_API_KEY}"
      }
    }
  }
}

What it does

Tavily builds search for AI agents, and its tavily-mcp package adds tools such as tavily-search and tavily-extract. The agent gets ranked results with snippets and can pull the readable content of specific URLs.

It starts locally with npx and reads your key from TAVILY_API_KEY. An optional DEFAULT_PARAMETERS variable, a JSON string, sets defaults like search depth and result count.

Example

you › Search for the current LTS version of Node and quote the release date.

agent › The newest LTS line is listed on nodejs.org's releases page with its start date and end-of-life date. The agent cites the source URL beside each figure.

Questions

Is there a hosted version?
Yes, Tavily also serves a remote endpoint at mcp.tavily.com that accepts OAuth or a Bearer key. This setup uses the local package.
How do I limit result counts?
Set DEFAULT_PARAMETERS to a JSON string such as {"max_results": 5, "search_depth": "basic"}.
  • Asana MCP Server

    MCP setupWorkflow

    Free
  • Atlassian MCP Server

    MCP setupWorkflow

    Free
  • Brave Search MCP Server

    MCP setupWorkflow

    Free
  • ClickUp MCP Server

    MCP setupWorkflow

    Free