The arXiv MCP server turns Claude into a research assistant for papers. It searches arXiv by topic, category and date, downloads papers as markdown, and reads them one section at a time, including the author's original LaTeX. It also follows citations both ways through Semantic Scholar and exports BibTeX from arXiv's own metadata.
It's a community project, runs locally and needs no API key. Use it when a feature, a model choice or an article depends on what the research actually says. For general web search, use Exa or Perplexity instead.
Watch out
- Papers on arXiv aren't peer reviewed by default. Ask the agent to say when a result comes from a preprint.
- Reads are capped at 12,000 characters per call, so long papers take several steps. That's intended: it keeps the context small.
- Semantic search across downloaded papers needs the optional
[pro]install.
Install
Terminal
Loading editor...
claude mcp add --transport stdio --scope user arxiv -- uvx arxiv-mcp-server
More in Top MCP servers for web data.