arXiv

Provides direct access to arXiv's scientific paper repository with tools for searching, retrieving metadata, downloading PDFs, and loading full article content into conversation context.

The arXiv MCP server provides AI assistants with direct access to arXiv.org's scientific paper repository through a set of specialized tools. Built by Prashal Ruchiranga, it enables searching, retrieving metadata, downloading PDFs, and loading full article content into context. The implementation uses feedparser to query the arXiv API and PyMuPDF to extract text from PDFs, with configurable download paths for storing retrieved documents. This server is particularly useful for research workflows requiring access to scientific literature, allowing AI systems to search by title, author, or abstract and work directly with academic content without leaving the conversation context.