MertAtesmen/tavily-mcp-sse

Tavily MCP Server

Tavily MCP Server implementation that uses fastmcp and supports both sse and stdio transports. To use this server, you need a Tavily account and a Tavily API key, which must be loaded into the TAVILY_API_KEY environment variable.

The Tavily MCP server provides:

• search, extract, map, crawl tools
• Real-time web search capabilities through the tavily-search tool
• Intelligent data extraction from web pages via the tavily-extract tool
• Powerful web mapping tool that creates a structured map of website
• Web crawler that systematically explores websites

Prerequisites

• git installed. (To clone the repo)
• uv installed.
• docker installed (Optional: If you are planning to use the SSE server inside a docker container).

To install uv in Linux and MacOS type this in your terminal:

curl -LsSf https://astral.sh/uv/install.sh | sh

Environment Variables

Copy the .env.example file and rename that to .env. Then paste your TAVILY_API_KEY inside there

TAVILY_API_KEY=<YOUR-API-KEY>

Optional: You can also configure the port if you are planning to use SSE.

TAVILY_MCP_PORT=<PORT>

Running the SSE server

While inside the repo run:

uv run --env-file .env tavily-mcp-sse

Running on STDIO

{
  "mcpServers": {
    "tavily-mcp-server": {
      "command": "uv",
      "args": [
        "run",
        "--directory",
        "<LOCATION-TO-THE-REPO>",
        "tavily-mcp-stdio"
      ],
      "env": {
        "TAVILY_API_KEY": "<YOUR-API-KEY>"
      }
    }
  }
}

Docker SSE Server

First you need to build the image using the Dockerfile inside this repository. Run this to build the image:

docker build -t tavily-mcp .

Then you can run the container using the environment variables inside the env file

docker run --name tavily-mcp \
  -p 127.0.0.1:8000:8000 \
  --env-file .env \
  tavily-mcp

Or you can specify the environment variables yourself.

docker run --name tavily-mcp \
  -p 127.0.0.1:8000:8000 \
  -e TAVILY_API_KEY=<YOUR-API-KEY>
  tavily-mcp