ArXiv MCP Server: An MCP Server for Searching and Analyzing arXiv Papers 157

What is the ArXiv MCP Server?

The ArXiv MCP Server is an MCP server that provides arXiv paper search and analysis capabilities for AI assistants. Through the Model Context Protocol (MCP) interface, it allows AI models to search, download, view, and read papers in a programmatic way, and also provides research prompts to assist with analysis. It supports local storage of papers for quick access and can be applied to scenarios such as academic research assistance and AI assistant integration.

ArXiv MCP Server Features

• Paper Search: Allows querying arXiv papers with precise filtering by date range, categories, etc., for example: result = await call_tool("search_papers", {"query": "transformer architecture", "max_results": 10, "date_from": "2023-01-01", "categories": ["cs.AI", "cs.LG"]}).

• Paper Download: Downloads papers by arXiv ID, such as result = await call_tool("download_paper", {"paper_id": "2401.12345"}).

• Paper List Viewing: Able to view all downloaded papers by calling result = await call_tool("list_papers", {}).

• Local Storage: Downloaded papers are saved locally for quick access; the storage location can be configured via the ARXIV_STORAGE_PATH environment variable, which defaults to ~/.arxiv-mcp-server/papers.

• Research Prompts: Provides dedicated prompts to help analyze academic papers, such as result = await call_prompt("deep-paper-analysis", {"paper_id": "2401.12345"}). This prompt includes detailed instructions, a systematic analysis workflow, and a comprehensive analysis structure covering executive summary, research background, method analysis, result evaluation, practical and theoretical implications, future research directions, and broader impact.

ArXiv MCP Server Use Cases

• Academic Research Assistance: Provides researchers with convenient channels for searching and obtaining papers, helping them quickly find the academic resources they need and saving time on literature retrieval, so they can carry out research work more efficiently.

• AI Assistant Integration: Acts as a bridge between AI assistants and the arXiv paper repository, enabling AI assistants to access and analyze arXiv papers through a simple MCP interface, providing users with richer knowledge and information support. For example, when answering academic questions, an AI assistant can quickly retrieve relevant papers and provide accurate information.

• Knowledge Management and Sharing: Facilitates team or individual management and sharing of papers. Through local storage and paper list viewing features, users can easily organize and review downloaded papers, promoting the accumulation and dissemination of knowledge.

ArXiv MCP Server Tutorial

Installation:

• Install via Smithery: npx -y @smithery/cli install arxiv-mcp-server --client claude.

• Manual Installation: Install using the uv tool, with the command uv tool install arxiv-mcp-server; for a development environment, you need to first clone the project code, create and activate a virtual environment, then install the test dependencies. The specific operations are:

git clone https://github.com/blazickjp/arxiv-mcp-server.git
cd arxiv-mcp-server
uv venv
source .venv/bin/activate
uv pip install -e ".[test]"

MCP Integration:

Add the following configuration to the MCP client configuration file:

{
"mcpServers": {
"arxiv-mcp-server": {
"command": "uv",
"args": [
"tool",
"run",
"arxiv-mcp-server",
"--storage-path", "/path/to/paper/storage"
]
}
}
}

For a development environment, the args in the configuration need to add "--directory", "path/to/cloned/arxiv-mcp-server".

Tool Usage:

Implement different functions by calling the corresponding tool functions, such as the search_papers, download_paper, list_papers, and read_paper mentioned above.

Testing:

Run the test suite to ensure functionality works correctly, with the command python -m pytest.

GitHub project: https://github.com/blazickjp/arxiv-mcp-server