Pubmed MCP Server
PubMed database searching and article metadata retrieval for AI models.
Data last scanned yesterday · Reviewed last month
Overview
PubMed MCP Server enables LLMs to interface with the NCBI E-utilities API. It allows AI assistants to perform complex literature searches, retrieve abstracts, and fetch comprehensive article metadata across the PubMed biomedical database. The server supports batch operations and multiple output formats including XML and JSON. It includes built-in rate limiting to manage request volume and supports optional NCBI API key configuration to increase throughput for large-scale data retrieval.
Our verdict
This server carries a high risk level with a security score of 10/100. The repository has not been updated for over a year (last commit 397 days ago) and lacks any legal license, making its usage terms unclear. Due to these maintenance and legal concerns, it should only be used in non-critical or experimental environments by users who have manually audited the codebase.
- Zero direct dependencies reduce supply-chain surface area
- Supports automatic rate limiting for API compliance
- Provides real, functional documentation in the README
- High risk due to 10/100 security score
- Stale repository with no commits for 397 days
- Missing license terms for legal usage
- Limited community presence with only one star
Tools
| Tool | Description | Risk |
|---|---|---|
| search_articles | Search PubMed for articles matching a query. | low |
| download_article | Download article details by PubMed ID. | low |
| download_articles_batch | Download multiple articles in a single request. | low |
| get_article_summaries | Get document summaries for articles (metadata without full content). | low |
Compatibility
| Client | Local | Docker | Remote | Read-only |
|---|---|---|---|---|
| Claude Desktop | ||||
| Cursor | ||||
| VS Code | ||||
| Windsurf | ||||
| ChatGPT |
Frequently asked questions
›What is the benefit of providing an NCBI API key?
Providing an NCBI API key increases your PubMed API rate limit from 3 to 10 requests per second.
›Can I search PubMed using specific fields?
Yes, you can use field-specific tags in your query, such as [Title], [Author], [Journal], [PDAT] for publication date, or [MeSH] for MeSH terms.
›What output formats are supported for article downloads?
The server supports downloading article content in XML, JSON, and text formats.
›Does the server support batch operations?
Yes, you can use the download_articles_batch tool to retrieve content for multiple PubMed IDs in a single request.
›How do I configure the server for use with Claude Desktop?
You can configure the server in your Claude Desktop configuration file by pointing to your Python executable and script, optionally setting your NCBI API key and email in the env block or via a .env file.
›What happens if I exceed the PubMed API rate limits?
The server includes automatic rate limiting functionality to help ensure compliance with PubMed's API usage policies.
Alternatives
Badge
Maintain this server? Add the live badge to your README.