arXiv MCP Server

A powerful Model Context Protocol (MCP) server that provides seamless access to arXiv, the world's largest open-access archive of scholarly articles. This server enables you to search, discover, and retrieve academic papers across multiple disciplines including physics, mathematics, computer science, and more. Whether you're conducting literature reviews, tracking research trends, or exploring cutting-edge scientific developments, this server brings arXiv's vast repository directly into your workflow.

Overview

The arXiv MCP server acts as a bridge between your applications and arXiv's extensive database of over 2 million scholarly papers. It provides sophisticated search capabilities with support for Boolean operators, field-specific queries, and category-based filtering. Beyond simple searches, the server offers structured access to paper metadata, abstracts, and PDF links, making it an essential tool for researchers, students, and anyone working with academic literature.

Features

Comprehensive Search Capabilities

Search arXiv's database using natural language queries or advanced search syntax. The server supports Boolean operators (AND, OR, ANDNOT) and field-specific searches, allowing you to precisely target your queries. Search within specific fields like titles, abstracts, author names, or category codes to find exactly what you need.

Author-Specific Queries

Track the work of specific researchers by searching for papers by author name. This feature is particularly useful for conducting comprehensive literature reviews, following influential researchers in your field, or exploring the publication history of potential collaborators.

Category-Based Discovery

Browse papers within specific arXiv categories to stay current with developments in particular subfields. With access to dozens of categories spanning computer science, physics, mathematics, biology, and more, you can efficiently monitor research areas relevant to your work.

Flexible Pagination and Sorting

Control result sets with customizable pagination options and multiple sorting criteria. Sort papers by relevance, submission date, or last updated date in ascending or descending order. This flexibility ensures you can organize results to match your specific research needs.

Direct Paper Access

Retrieve detailed information about specific papers using arXiv IDs, including complete metadata, abstracts, and direct PDF download links. This structured access eliminates the need to manually navigate the arXiv website while maintaining all the information you need.

Tools

search_papers

Search for academic papers using a general query with advanced search capabilities.

Parameters:

• query (required, string): Search query string. Supports Boolean operators and field prefixes like ti: for title, au: for author, abs: for abstract, and cat: for category. Examples: "machine learning", "ti:transformer AND cat:cs.LG", "au:Hinton"
• max_results (optional, number, default: 10): Maximum number of results to return
• start (optional, number, default: 0): Starting index for pagination
• sort_by (optional, string, default: "relevance"): Sort order - choose from "relevance", "lastUpdatedDate", or "submittedDate"
• sort_order (optional, string, default: "descending"): Sort direction - either "ascending" or "descending"

search_papers_by_author

Search for papers by a specific author name.

Parameters:

• author (required, string): Author name, such as "Geoffrey Hinton" or "Yann LeCun"
• max_results (optional, number, default: 10): Maximum number of results to return
• start (optional, number, default: 0): Starting index for pagination

search_papers_by_category

Search for papers within a specific arXiv category.

Parameters:

• category (required, string): arXiv category code such as "cs.AI" for Artificial Intelligence, "cs.LG" for Machine Learning, or "math.CO" for Combinatorics. Use the get_categories tool to see all available categories
• max_results (optional, number, default: 10): Maximum number of results to return
• start (optional, number, default: 0): Starting index for pagination

get_categories

Retrieve a complete list of available arXiv category codes and their descriptions.

Parameters: None required. This tool returns the full taxonomy of arXiv categories to help you construct targeted searches.

Resource Templates

paper

Retrieve comprehensive information about a specific paper using its arXiv ID.

URI Format: arxiv://paper/{arxiv_id}

Example: arxiv://paper/2301.07041 or arxiv://paper/1706.03762v7

Returns complete paper metadata including title, authors, abstract, publication dates, categories, and links.

paper_abstract

Retrieve only the abstract of a specific paper.

URI Format: arxiv://paper/{arxiv_id}/abstract

Example: arxiv://paper/2301.07041/abstract

Returns just the abstract text, useful when you need a quick summary without the full metadata.

paper_pdf

Get the direct PDF download URL for a specific paper.

URI Format: arxiv://paper/{arxiv_id}/pdf

Example: arxiv://paper/2301.07041/pdf

Returns the URL to download the paper's PDF, enabling direct access to the full text.

Use Cases

Literature Reviews

Conduct comprehensive literature reviews by searching across multiple dimensions - by keyword, author, category, and time period. The flexible search tools allow you to cast a wide net initially, then refine your results using advanced Boolean operators and field-specific searches.

Research Monitoring

Stay current with the latest research in your field by regularly querying specific categories or following prolific authors. The sorting options ensure you see the most recent papers first, helping you track emerging trends and breakthrough developments.

Citation and Reference Management

Quickly retrieve paper details using arXiv IDs when building bibliographies or verifying references. The structured metadata format makes it easy to extract citation information and integrate it into reference management workflows.

Cross-Disciplinary Exploration

Discover connections between different fields by exploring multiple categories and searching for papers that bridge disciplines. The comprehensive category system makes it easy to venture beyond your primary area of expertise.

Prompts

This server does not currently define any pre-configured prompts.