PubMed

You can search PubMed, fetch detailed article information, and retrieve full-text data (when available) through an MCP server built for PubMed. This server handles query parsing, data retrieval, and optional caching to improve response times, all while respecting PubMed’s rate limits.

How to use

You interact with the PubMed MCP server through an MCP client. You can perform the following actions to access PubMed data: search for articles using flexible query parameters; fetch detailed summaries for specific PubMed IDs; and retrieve full-text information when it is available.

Available tools and their purposes:

• search — find PubMed articles by a query string, with optional max_results and sort order
• fetch_summary — get detailed information for a list of PubMed IDs (PMIDs)
• get_full_text — obtain full-text data for PMIDs when provided by PubMed or partners

How to install

Prerequisites you need before installing the PubMed MCP server:

• Node.js 18 or later installed on your machine

• npm (comes with Node.js) or a compatible package manager

Choose whether to install the MCP server for project scope (local to your project) or globally.

Install in project scope to keep the MCP server configuration inside your project directory.

# Project-scoped installation (creates a .mcp.json in your project root)
claude mcp add pubmed-mcp \
  --scope project \
  npx -y @ncukondo/pubmed-mcp \
  --email your@email.com

# Separate line for explicit installation command if you omit the project scope option
claude mcp add pubmed-mcp \
  npx -y @ncukondo/pubmed-mcp \
  --email your@email.com

Additional setup tips

If you want to enable higher rate limits, you can provide an API key from PubMed. You can either pass the API key in the command (not shown in every example) or set the environment variable PUBMED_API_KEY.

Available tools

search

Search PubMed articles with a flexible set of query parameters, returning a list of matching articles.

fetch_summary

Fetch detailed summaries for a list of PubMed IDs, including abstracts, authors, and DOIs.

get_full_text

Retrieve available full-text information for PubMed articles.