Tavily
A comprehensive search API for real-time web search, data extraction, and crawling, requiring a Tavily API key.
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
Related Servers
PulseMCP Server
Discover and explore MCP servers and integrations using the PulseMCP API.
Tavily Search
A search engine powered by the Tavily AI Search API.
CoolPC MCP Server
Query computer component prices from Taiwan's CoolPC website to generate AI-assisted price quotes.
Open Custom Search API
Perform web searches using Google's Custom Search API.
NRTSearch
Exposes Lucene-based search indexes to AI assistants through the NRTSearch MCP server.
Baidu Search
A search server for the Model Context Protocol (MCP) that uses the Baidu Wenxin API.
arXiv Research Assistant
Interact with the arXiv.org paper database. Supports keyword search, paper lookups, author searches, and trend analysis.
Jina AI Search
Access Jina AI's Search Foundation APIs for web search, news search, and more, tailored for LLMs.
OpenStreetMap
Enhances LLMs with location-based services and geospatial data from OpenStreetMap.
Gemini DeepSearch MCP
An automated research agent using Google Gemini models and Google Search to perform deep, multi-step web research.