bioRxiv MCP Server

The bioRxiv MCP Server is a community-built Model Context Protocol server that exposes the bioRxiv preprint repository to AI assistants. It enables LLMs to perform keyword searches, advanced multi-parameter queries, and DOI-based metadata lookups against bioRxiv, the open-access archive for unpublished biology preprints.

Built on FastMCP in Python, the server provides three core tools: a simple keyword search, an advanced search supporting structured parameters, and a metadata fetcher that returns full record details for a given DOI. There is no authentication required because bioRxiv exposes a public API.

This is a community implementation, not an official server from bioRxiv. A companion project by the same author, medRxiv-MCP-Server, provides equivalent access to the medRxiv health sciences preprint archive. Together they cover the bioRxiv and medRxiv ecosystems, but they must be installed as two separate MCP servers.

Tools

Prerequisites

• Python 3.10 or higher
• An MCP-capable client (Claude Desktop, Cursor, etc.)
• No API key required (bioRxiv API is public)

Option 1: Install via Smithery

npx -y @smithery/cli@latest install @JackKuo666/biorxiv-mcp-server --client claude --config "{}"

Option 2: Manual install

git clone https://github.com/JackKuo666/bioRxiv-MCP-Server.git
cd bioRxiv-MCP-Server
pip install -r requirements.txt

Claude Desktop config (macOS / Linux)

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "biorxiv": {
      "command": "python",
      "args": ["-m", "biorxiv-mcp-server"]
    }
  }
}

Claude Desktop config (Windows)

{
  "mcpServers": {
    "biorxiv": {
      "command": "C:\\Users\\YOUR_USERNAME\\AppData\\Local\\Programs\\Python\\Python311\\python.exe",
      "args": ["-m", "biorxiv-mcp-server"]
    }
  }
}

For medRxiv access, install the companion server at https://github.com/JackKuo666/medRxiv-MCP-Server separately.

• Pull the latest bioRxiv preprints on a topic (e.g. CRISPR base editing) into a literature review draft.
• Resolve a DOI mentioned in a chat conversation into full preprint metadata including authors, abstract, and version history.
• Run an advanced search filtered by subject collection (e.g. "neuroscience") and a date range to monitor newly posted research.
• Build an automated daily digest of new bioRxiv papers matching specific keywords for a research lab.
• Cross-reference a bioRxiv preprint's metadata against citation databases inside an agent workflow.

• "Search bioRxiv for recent preprints about single cell RNA sequencing in zebrafish."
• "Find bioRxiv papers tagged neuroscience posted between 2026-01-01 and 2026-04-30 mentioning Alzheimer's."
• "Get the full metadata for bioRxiv DOI 10.1101/2024.05.12.593789."
• "List the 10 most recent bioRxiv preprints on protein structure prediction."
• "Summarize the abstracts of the top 5 bioRxiv hits for 'mRNA vaccine adjuvant'."

• No API key or authentication required; works out of the box against the public bioRxiv API.
• Exposes both simple keyword search and a structured advanced search with date and subject filters.
• Small, focused tool surface (three tools) makes it easy for an LLM to use reliably.
• A matching medRxiv server from the same author covers the medical preprint sibling archive.

• Community-maintained, not an official project from bioRxiv or Cold Spring Harbor Laboratory.
• Scoped to bioRxiv only; medRxiv requires installing a separate MCP server.
• Limited to search and metadata retrieval, no built-in full-text PDF parsing or citation graph features.

• JackKuo666/medRxiv-MCP-Server for medRxiv health sciences preprints.
• benedict2310/Scientific-Papers-MCP for a unified server covering arXiv, OpenAlex, PMC, Europe PMC, bioRxiv/medRxiv, and CORE.
• arijitnist/Biomedical-MCP-Servers bundles multiple biomedical sources including bioRxiv and medRxiv.