Overview

The arXiv MCP Server provides AI assistants with access to one of the world's most important open-access research repositories. With over 2 million scholarly papers across physics, mathematics, computer science, quantitative biology, finance, and statistics, arXiv is the primary platform for sharing cutting-edge research. This server enables AI models to search papers, read abstracts, access metadata, and retrieve LaTeX source.

Community MCP Server:

Integrates with the official arXiv API for paper search and metadata retrieval

Key Features

🔍

Paper Search

Search across 2M+ papers by keyword, author, title, or category with advanced query syntax

📋

Abstract Access

Read full paper abstracts to quickly understand research contributions and findings

🏷️

Category Navigation

Browse papers by arXiv category including cs.AI, math.OC, physics, and more

👤

Author Search

Find all papers by a specific author with publication history and co-authors

📄

LaTeX Source

Access LaTeX source code for papers with mathematical and technical content

📊

Paper Metadata

Get complete metadata including publication date, categories, DOI, and citation info

Available Tools

Quick Reference

Tool	Purpose
`search_papers`	Search arXiv papers by keyword, author, or title
`get_paper_details`	Get full metadata and abstract for a specific paper
`list_categories`	List all arXiv categories and subcategories
`get_author_papers`	Find all papers by a specific author
`get_recent_papers`	Get recently published papers in a category

Detailed Usage

search_papers▶

Search arXiv papers with flexible query options across all fields.

{
  "query": "transformer attention mechanism",
  "category": "cs.LG",
  "max_results": 20,
  "sort_by": "relevance"
}

get_paper_details▶

Get detailed metadata and abstract for a specific paper by its arXiv ID.

{
  "paper_id": "2301.12345"
}

get_author_papers▶

Find all papers by a specific author with sorting and filtering options.

{
  "author_name": "Andrej Karpathy",
  "max_results": 50
}

get_recent_papers▶

Get the most recently published papers in a specific arXiv category.

{
  "category": "cs.AI",
  "max_results": 30
}

Installation

{
  "mcpServers": {
    "arxiv": {
      "command": "npx",
      "args": ["-y", "arxiv-mcp"]
    }
  }
}

No API Key Required:

Uses the public arXiv API — no authentication required for searching and accessing paper metadata

Use Cases

Literature Discovery

Ask AI to find the most relevant papers on any research topic with detailed summaries.

Research Tracking

Monitor new papers from specific authors or categories to stay current in your field.

Literature Review

Gather and summarize papers for comprehensive literature reviews on any topic.

Citation Analysis

Find highly cited papers, track research trends, and identify influential works in a field.

Sources

DataBridge MCP Server

DataBridge MCP servers enable AI models to interact with local databases for contextual information, supporting persistent storage and unified access to ML services.

AI Research Assistant MCP Server

AI Research Assistant MCP server provides autonomous deep research capabilities — searching, analyzing, and synthesizing information from across the web into comprehensive research reports.

Canva Dev MCP Server

The Canva Dev MCP Server enables AI models to interact with the Canva development ecosystem, providing capabilities for generating apps, adopting the App UI Kit, localizing apps, and ensuring compliance with Canva's design guidelines.

arXiv MCP Server