OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
MCP & Tooling

Arxiv MCP Server

Arxiv MCP Server is an open-source Python MCP server that connects compatible AI clients to tools for arXiv literature work. It supports searching and retrieving paper information, reading paper content and original-LaTeX sections, generating BibTeX from arXiv metadata, and monitoring topics. Paper files and related data are stored locally by default, while some discovery and citation tools call external services.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Arxiv MCP Server — OpenViST project profile in MCP & Tooling.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when an MCP-compatible AI client needs tools for arXiv research, especially a workflow that moves from paper discovery or an ID to an outline, selected section and citation. Search is optional for workflows that begin with a known paper.

Who it is for

  • Researchers who use MCP-compatible AI clients to explore arXiv papers.
  • Developers connecting an agent to paper-reading, citation and topic-monitoring tools.
  • Users who want downloaded paper files retained in local storage.

Documented uses

  • Find papers by query, category, date or sort order, or start with a known arXiv ID.
  • Inspect a paper outline and retrieve a selected section, including from author-submitted LaTeX.
  • Create BibTeX citations from arXiv metadata.
  • Keep topic watches and paper files on disk for a continuing literature workflow.
Before you choose

Requirements & limits

The documented default is a local stdio server launched with uvx; uv is required for that command. The project is open source under Apache-2.0. A Semantic Scholar API key is available to reduce citation-graph rate-limit concerns; unauthenticated requests work until quota is exhausted.

  • The default paper directory is ~/.arxiv-mcp-server/papers; a different absolute path can be set with --storage-path.
  • The default transport is stdio, and client configuration formats vary.
  • Search and source retrieval call external services even though the server and paper storage run locally.
  • Citation-graph lookup uses Semantic Scholar and may be rate-limited without an API key.
  • The supported package is on PyPI; an unrelated npm package has the same name.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install the server for an MCP client

Launch the server through uvx and add its stdio configuration to a compatible client.

Before you start

  • uv installed, providing the uvx command.
  • An MCP client that accepts the documented mcpServers JSON shape, or client-specific MCP configuration.

Steps

  1. Run uvx arxiv-mcp-server as the documented default launch command.

  2. For clients that accept the mcpServers JSON shape, add an entry with type stdio, command uvx, and args containing arxiv-mcp-server.

  3. Use the client's own MCP documentation if it expects a different configuration format.

Expected outcome

The client is configured to launch the local arXiv MCP server over stdio.

Guide limitations

  • The JSON shape shown is not universal; clients may use a top-level servers object, TOML or a settings interface.
  • The supported package is published on PyPI; the repository warns against installing the unrelated same-name npm package.
Documentation 1 ↗
Source-based practical guide

Choose a local paper storage directory

Set the directory used for paper and related local storage in the server configuration.

Before you start

  • An MCP client configuration that accepts command arguments.

Steps

  1. Use the default paper directory ~/.arxiv-mcp-server/papers, or choose an absolute path.

  2. To use a different directory, append --storage-path and the absolute path as arguments after arxiv-mcp-server in the client configuration.

Expected outcome

The server is configured to use the selected local storage path.

Guide limitations

  • This configures storage; it does not itself download or read a paper.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
10 October 2026
Original website title
blazickjp/arxiv-mcp-server · Official homepage title · retained in its source language
How we review sourcesSuggest a correction