Unofficial PubMed

Provides MCP access to PubMed with keyword search, advanced filtering, metadata, and PDF retrieval.

javascript

0

GitHub Stars

javascript

Language

6 months ago

First Indexed

3 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "openpharma-org-pubmed-mcp": {
      "command": "node",
      "args": [
        "/path/to/pubmed-mcp-server/build/index.js"
      ]
    }
  }
}

You can access PubMed’s vast biomedical literature through an MCP server that exposes keyword and advanced search, detailed article metadata, and PDF access where available. This enables AI assistants and applications to search the PubMed database programmatically, retrieve rich article information, and fetch full-text PDFs via MCP clients you trust.

How to use

Connect your MCP client to the PubMed MCP server you’ve configured. Use the available tools to search for articles, refine results with advanced criteria, fetch detailed metadata for a specific PubMed ID (PMID), or download PDFs when available. You can combine queries to target diseases, interventions, authors, or journals, and you’ll receive structured results that include titles, authors, journals, publication dates, abstracts, DOIs, PMCID, MeSH terms, keywords, and URLs.

How to install

Prerequisites: ensure Node.js is installed on your system.

# (1) Ensure you have a Node.js runtime
node -v
npm -v

# (2) Prepare the PubMed MCP Server entry point (as shown in the example configuration)
# The server is started by running the Node script at the given path
# You may replace the path with the actual build output location
node /path/to/pubmed-mcp-server/build/index.js

Additional sections

Configuration is provided as a ready-to-use MCP server entry. The following example shows how to register the server with an MCP client. It uses a standard stdio approach where Node runs the executable script that powers the server.

{
  "mcpServers": {
    "pubmed": {
      "command": "node",
      "args": ["/path/to/pubmed-mcp-server/build/index.js"]
    }
  }
}

Notes and capabilities

This server provides four primary capabilities: search_keywords for keyword-based queries, search_advanced for filtered and date-bounded searches, get_article_metadata for detailed article information, and get_article_pdf to attempt downloading the article’s full-text PDF when available. Results include key bibliographic fields such as pmid, title, authors, journal, publication date, abstract, doi, pmcid, keywords, MeSH terms, and a stable URL for reference.

Available tools

search_keywords

Search PubMed using keywords and natural language queries.

search_advanced

Perform filtered searches with specific criteria such as term, title, author, journal, and date range.

get_article_metadata

Retrieve detailed metadata for a specific PubMed ID (PMID) including title, authors, abstract, and identifiers.

get_article_pdf

Attempt to access the full-text PDF of an article when available via PMC.