What it does

Academia MCP searches academic paper repositories (arXiv, ACL Anthology, Hugging Face datasets) and retrieves citation networks via Semantic Scholar. It fetches and normalizes papers as structured text, compiles LaTeX documents, extracts PDF content, and optionally provides LLM-powered tools for document analysis and research proposal generation. Web search integration via Exa, Brave, or Tavily adds real-time context to paper research.

Who it's for

Researchers, PhD candidates, and academic writers conducting literature reviews. Anyone building a synthesis workflow that combines paper discovery, citation tracking, and contextual web search. Teams using Claude to ground analysis in peer-reviewed sources.

Common use cases

• Query arXiv or ACL Anthology and download matching papers as structured text
• Trace citation graphs backward (references) or forward (papers citing a work)
• Search Hugging Face datasets to find data relevant to a research question
• Web search for recent blog posts, news, or implementation details related to a paper
• Batch-process paper collections with LLM to generate summaries or answer specific questions

Setup pitfalls

• Python 3.12+ is required; older environments will fail to install.
• Web search requires at least one of EXA_API_KEY, BRAVE_API_KEY, or TAVILY_API_KEY set; without them, web search tools become unavailable.
• LLM-powered tools require OPENROUTER_API_KEY; without it, document QA and research proposal generation are unavailable.
• PDF extraction and LaTeX compilation require WORKSPACE_DIR to be set to a writable directory; missing this disables those features.