PubMed

The PubMed MCP Server provides a bridge between AI assistants and PubMed's biomedical literature. It exposes a set of MCP tools that let you search for PubMed articles, fetch article metadata, download PDFs when available, and perform deep analyses of papers. You can integrate it with AI desktop clients like Claude, Cursor, Windsurf, and Cline by pointing the MCP client at the server using the appropriate command and arguments shown in the README. Common workflows include searching for papers with keywords, retrieving detailed metadata for a given PMID, and then requesting a deep analysis of a paper.

Core tools available:

• search_pubmed_key_words: search PubMed using keyword queries.
• search_pubmed_advanced: perform advanced searches with multiple fields and filters.
• get_pubmed_article_metadata: fetch metadata for a paper by PMID.
• download_pubmed_pdf: attempt to download the full-text PDF when accessible.
• deep_paper_analysis: run a comprehensive analysis of a paper's content.

To use with Claude Desktop or other MCP clients, configure the client to run the server command (python -m pubmed-mcp-server) and connect to the server as an MCP source. You can then issue natural language prompts that invoke these tools, such as asking for recent CRISPR papers or requesting a metadata overview for a specific PMID. The server is designed to be lightweight and suitable for research workflows, enabling rapid integration of PubMed data into AI-assisted research pipelines.

Prerequisites

• Python 3.10+
• internet access to install dependencies from PyPI
• pip (Python package installer)

Step-by-step installation

1. Clone the repository

git clone https://github.com/JackKuo666/PubMed-MCP-Server.git
cd PubMed-MCP-Server

2. Create and activate a Python environment (optional but recommended)

python -m venv venv
# macOS/Linux
source venv/bin/activate
# Windows
venv\Scripts\activate

3. Install required dependencies

pip install -r requirements.txt

4. Run the MCP server

python pubmed_server.py

5. Verify the server is running by connecting an MCP client to the configured server name (e.g., pubmed) using the provided command structure.

Tips and remarks:

• This server uses Python 3.10+ and depends on the FastMCP framework. Ensure you have network access to PubMed and that you comply with PubMed's terms of service.
• The README suggests optional client integrations (Claude, Cursor, Windsurf, Cline). When configuring clients, you will typically point the client to run: python -m pubmed-mcp-server and provide the appropriate mcpServers configuration on the client side.
• If you encounter missing dependencies, re-run pip install -r requirements.txt or install additional packages as needed.
• The environment dictionary under the mcpServers entry is provided for future variables; populate as needed (e.g., API keys, proxy settings).