Simple arXiv MCP Server

Data Science ToolsPython

Access and search scientific papers from arXiv

Available Tools

search_papers

Search for scientific papers on arXiv by title and abstract content

get_paper_details

Retrieve detailed information about a specific paper including abstract, authors, and links to available formats

Simple arXiv provides LLMs with the ability to search and retrieve scientific papers from arXiv's extensive database. This tool enables AI assistants to find relevant research papers by keyword, retrieve abstracts and metadata, and provide links to full papers in various formats. The server implements proper rate limiting according to arXiv's API guidelines, ensuring responsible usage of the service. With Simple arXiv, AI assistants can help users stay up-to-date with the latest research in their field without leaving the conversation.

Installation

Using Smithery (Recommended)

The easiest way to install Simple arXiv is through Smithery:

npx -y @smithery/cli install mcp-simple-arxiv --client claude

Manual Installation

You can also install the package manually using pip:

pip install mcp-simple-arxiv

Configuration

After installation, you'll need to configure your AI client to use the Simple arXiv MCP server.

For Claude Desktop

Add the following configuration to your claude_desktop_config.json file:

macOS:

{
  "mcpServers": {
    "simple-arxiv": {
      "command": "python",
      "args": ["-m", "mcp_simple_arxiv"]
    }
  }
}

Windows:

{
  "mcpServers": {
    "simple-arxiv": {
      "command": "C:\\Users\\YOUR_USERNAME\\AppData\\Local\\Programs\\Python\\Python311\\python.exe",
      "args": [
        "-m",
        "mcp_simple_arxiv"
      ]
    }
  }
}

After adding the configuration, restart Claude Desktop for the changes to take effect.

Usage

Once configured, you can use Simple arXiv through natural language requests to your AI assistant. Here are some examples:

Searching for Papers

You can search for papers by asking questions like:

"Can you search arXiv for recent papers about large language models?"
"Find papers on quantum computing published in the last year"
"Search for research on climate change mitigation strategies"

Getting Paper Details

Once you have a paper ID (usually in the format YYMM.NNNNN), you can request more details:

"Can you show me the details for paper 2103.08220?"
"Get the abstract for arXiv paper 2201.03545"
"What's the full information on that last paper you mentioned?"

The response will include the paper's title, authors, publication date, abstract, and links to available formats (PDF/HTML).

Features

Search for scientific papers by keywords
Filter search results by categories and date ranges
Retrieve paper metadata and abstracts
Access links to available paper formats (PDF/HTML)
Rate-limited according to arXiv's API guidelines (max 1 request every 3 seconds)

Troubleshooting

If you encounter issues:

Ensure Python is properly installed and in your PATH
Verify that the mcp-simple-arxiv package is installed (pip list | grep mcp-simple-arxiv)
Check that your client configuration is correct
Restart your AI client after making configuration changes