Arxiv Latex MCP logo

Arxiv Latex MCP

by takashiishida

Fetches and structures arXiv LaTeX so AI agents can analyze equations, sections, abstracts, and paper metadata.

Install Arxiv Latex MCP

Pick your client, then add it in one click or paste the config.

STDIOPython
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "takashiishida-arxiv-latex-mcp": {
      "command": "uvx",
      "args": [
        "arxiv-latex-mcp"
      ],
      "env": {
        "ARXIV_LATEX_MCP_CACHE": "YOUR_VALUE_HERE",
        "ARXIV_LATEX_MCP_CACHE_TTL_DAYS": "YOUR_VALUE_HERE",
        "ARXIV_LATEX_MCP_CACHE_DIR": "YOUR_VALUE_HERE"
      }
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

Values are written into the snippet above so you can paste it as is. They never leave your browser.

We couldn’t confirm it starts correctly

The install command started, but didn't respond the way we expected when we tried to talk to it.

uvx arxiv-latex-mcp

initialize succeeded but no response to tools/list.

This is an experimental automated check and can have false negatives — missing environment variables, a slow cold install, etc. It doesn’t necessarily mean something’s wrong. Last checked .

Overview

takashiishida/arxiv-latex-mcp gives MCP clients access to arXiv papers through their LaTeX source rather than PDFs. It can search papers, retrieve flattened source or abstracts, list section headings, and fetch individual sections. Use it for equation-heavy research where preserving mathematical notation matters, especially in computer science, mathematics, and engineering. The server uses arxiv-to-prompt for downloading, processing, and caching paper sources.

Use cases

  • Search arXiv for papers by topic, author, category, or date
  • Explain equations and notation from a paper's LaTeX source
  • Retrieve a paper's abstract or selected section
  • Review mathematical or technical papers with structured context

Key features

  • Fetches flattened LaTeX from arXiv papers
  • Searches papers with category and date filters
  • Retrieves abstracts and titled sections
  • Caches downloaded paper sources
  • Supports versioned and unversioned arXiv IDs

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked

Recent health check succeeded.

32/33 checks healthy over the last 103d

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 4h ago via OSV.dev · arxiv-latex-mcp (PyPI)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars151Stargazers on the source repository.
Last commit29d agoMost recent push to the default branch.
Install checkInconclusiveStarted but did not complete the MCP handshake in our test.
Directory activity3 viewsConfig copies, upvotes, and views on AllMCPs.

How Arxiv Latex MCP works

What takashiishida/arxiv-latex-mcp does

takashiishida/arxiv-latex-mcp connects MCP-compatible clients to arXiv paper content. Its main purpose is to expose the original LaTeX source in a form that language models can process, which helps retain equations, symbols, and notation that can be difficult to interpret from a PDF alone.

The server also supports paper discovery and structured reading. Agents can search by topic, title, or author, optionally constrain results by arXiv category and submission date, then use a returned paper ID to retrieve content. This makes it suitable for research assistants, literature review workflows, and questions about mathematical details in technical papers.

How it works

The server relies on arxiv-to-prompt to download and process arXiv LaTeX sources. A typical workflow begins with search_papers when the agent does not have an arXiv identifier. Search results include IDs, titles, authors, abstracts, categories, submission dates, and PDF links. The agent can then pass an ID to one of the paper-reading tools.

Paper sources are stored in arxiv-to-prompt's cache. Unversioned identifiers are refreshed after the configured cache lifetime so revised papers can be detected, while versioned identifiers remain fixed. If a refresh attempt fails, the cached copy is used. Search requests share a three-second throttle, and searches return up to 50 papers, with 10 results by default.

Setup and configuration

Claude Desktop users can install the downloadable .mcpb Desktop Extension. The extension supports macOS, Windows, and Linux, requires Python 3.10 or newer with pip, and downloads its dependencies into a private cache on first launch or when its runtime requirements change. Internet access is required for that initial dependency setup.

Other MCP clients can launch the package with uvx arxiv-latex-mcp. It can also be installed from PyPI with pip install arxiv-latex-mcp and started with arxiv-latex-mcp or python -m arxiv_latex_mcp. A local checkout can be run through uv with its directory set to the repository path. Restart the MCP client after changing its configuration or installing an updated extension.

Caching is controlled through MCP environment settings:

  • ARXIV_LATEX_MCP_CACHE: set to 0 to disable caching; defaults to enabled.
  • ARXIV_LATEX_MCP_CACHE_TTL_DAYS: controls when unversioned papers are refreshed; the default is 7 days.
  • ARXIV_LATEX_MCP_CACHE_DIR: selects the cache directory. Without it, arxiv-to-prompt chooses the platform default.

Tools and capabilities

  • search_papers finds papers by query, category, and inclusive UTC submission-date range, with relevance or newest-date sorting.
  • get_paper_prompt returns the paper's flattened LaTeX source.
  • get_paper_abstract returns only the abstract.
  • list_paper_sections returns the paper's section headings.
  • get_paper_section retrieves a section by title, including nested paths such as Section > Subsection.

Limitations and notes

The server fetches material from arXiv, so paper retrieval and first-time dependency setup require internet access. Search calls are throttled to one request interval of three seconds, and rate-limit responses require waiting before retrying. Cached content can be reused when refreshing a paper fails. Versioned arXiv IDs are treated as immutable references, while unversioned IDs may be downloaded again after the cache lifetime.

Written by AllMCPs with AI assistance from the project’s README and listing data. How we build listings

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Arxiv Latex MCP

takashiishida/arxiv-latex-mcp is an MCP server that connects AI clients to arXiv papers through their LaTeX sources. Its main tools search papers, return flattened LaTeX, fetch abstracts, list section headings, and retrieve specific sections, helping agents handle equations and mathematical notation more accurately than PDF-based workflows.

Install it with `uvx arxiv-latex-mcp`, or use `pip install arxiv-latex-mcp` and run `arxiv-latex-mcp` or `python -m arxiv_latex_mcp`. Claude Desktop users can install the downloadable `.mcpb` extension, which requires Python 3.10+ with pip.

Yes, Cursor is named as a supported MCP client. Configure it to launch the server with `uvx`, a pip-installed executable, or a local `uv` checkout, then restart the application after saving the MCP configuration.

Call `search_papers` with a query, optionally supplying categories, inclusive UTC date bounds, a result limit, and a relevance or newest-date sort. Use a returned paper ID with `get_paper_prompt` or the other reading tools.

Related MCP servers

Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewArxiv Latex MCP AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/takashiishida-arxiv-latex-mcp?style=directory)](https://allmcps.com/mcp/takashiishida-arxiv-latex-mcp)
HTML Embed
<a href="https://allmcps.com/mcp/takashiishida-arxiv-latex-mcp"><img src="https://allmcps.com/api/badge/takashiishida-arxiv-latex-mcp?style=directory" alt="Arxiv Latex MCP on AllMCPs" /></a>