Chat with AI and run tools instantly.
Search arXiv papers, retrieve metadata, read PDFs, and build literature review workflows from chat.
Live probe refreshed Sep 14, 2026 · Endpoint host mcp.mcpbundles.com
Built for
Academic Research, AI Papers, Literature Reviews, Diligence & Monitoring
Find recent papers
Uses query syntax, categories, and metadata for research discovery.
Search arXiv for recent papers on retrieval augmented generation in cs.CL and cs.AI, then group the strongest matches by method.
Read a paper deeply
Connect arXiv to any MCP client in minutes
Opens ChatGPT on the web or desktop and asks it to use the WebMCP tools available here.
You’ll sign in to MCPBundles when your client connects.
https://mcp.mcpbundles.com/bundle/arxivSign in once, then chat here with saved access — one connection to many servers, with a history of what your AI ran.
Chat with AI and run tools instantly.
Browse all toolsUses full-paper reading after metadata lookup.
Read this arXiv paper PDF and summarize the problem, method, datasets, results, limitations, and follow-up questions.
Build a literature review
Turns search results into structured review material.
Search arXiv for papers about AI code agents since 2024, fetch metadata for the top candidates, and organize them into a literature review outline.
Monitor a category
Supports recurring research monitoring workflows.
Find new cs.LG and stat.ML papers from the last month about model evaluation, then list titles, authors, abstracts, and PDF links.
What does the arXiv server provide?
It provides paper search, single-paper metadata lookup, batch metadata lookup, and full PDF reading for papers available through arXiv.
What search syntax should I use?
Use arXiv field prefixes such as all:, ti:, au:, abs:, cat:, and submittedDate ranges, combined with AND, OR, and ANDNOT for precise queries.
Can agents read full papers?
Yes. The full-paper tool returns the PDF for supported arXiv papers so agents can summarize methods, results, limitations, and citations in context.
Related editorial
arXiv Research Workflows with AI
How AI agents can run advanced arXiv searches, fetch paper metadata by ID, pull full PDFs, and assemble literature reviews on a topic.
Domain knowledge for arXiv — workflow patterns, data models, and gotchas for your AI agent.
Search, retrieve, and read scientific papers from arXiv.
Arxiv Get Paper
Get a specific arXiv paper by arXiv ID using the official arXiv API (Atom feed).
Arxiv Get Papers
Batch fetch multiple arXiv papers by ID in one call. Pass comma-separated arXiv IDs. Returns metadata for all requested papers. Much faster than individual get_paper calls due to arXiv's 3-second rate limit between requests.
Arxiv Read Paper
Read the full content of an arXiv paper. Returns the paper as a PDF document (inline embedded resource when under ~512 KB, otherwise an S3 presigned URL via resource_link). Use after search or single-paper metadata lookup when you need the full text, not just the abstract.
Arxiv Search Papers
Search arXiv for papers using the official arXiv API (Atom feed). Returns metadata including title, authors, abstract, categories, arXiv URL, PDF URL, and a normalized arxiv_id suitable for follow-up calls.
It provides paper search, single-paper metadata lookup, batch metadata lookup, and full PDF reading for papers available through arXiv.
Use arXiv field prefixes such as all:, ti:, au:, abs:, cat:, and submittedDate ranges, combined with AND, OR, and ANDNOT for precise queries.
Yes. The full-paper tool returns the PDF for supported arXiv papers so agents can summarize methods, results, limitations, and citations in context.
arXiv asks clients to avoid rapid sequential requests. Batch metadata lookups and scoped searches are better than many single-paper calls.
Add the MCPBundles server URL to your MCP client configuration (Claude Desktop, Cursor, VS Code, etc.). The URL format is: https://mcp.mcpbundles.com/bundle/arxiv. Authentication is handled automatically.
arXiv provides 4 tools that can be called by AI agents, along with a SKILL.md that gives your AI agent domain knowledge about when and how to use them.
arXiv uses open data APIs — no authentication required.
© 2026 ThinkChain Inc. All rights reserved.
MCPBundles is an independent platform built on the open Model Context Protocol standard. Not affiliated with Anthropic PBC or Claude.
Other MCP servers in this category from the directory index
Exa offers fast and intelligent web search capabilities, along with web crawling features to gather ...
Discover and book the best experiences worldwide! Immerse yourself in concerts, art, events, top att...
SuggestAPI exposes commerce tools through MCP and WebMCP on its Agent Gateway
Recall-first web search API for comprehensive event retrieval — finds every relevant event across th...
Real-time web search API and MCP for AI agents and RAG apps.
Unpaywall API providing open access status and full-text availability for scholarly articles.
5 tools