Paper Reader
An MCP server that empowers AI assistants (Cursor, Claude, Cherry Studio) to search, read, and summarize arXiv papers directly.
What is Paper Reader?
Paper Reader is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to mcp server that empowers ai assistants (cursor, claude, cherry studio) to search, read, and summarize arxiv papers directly.
An MCP server that empowers AI assistants (Cursor, Claude, Cherry Studio) to search, read, and summarize arXiv papers directly.
This server falls under the Search & Data Extraction category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- An MCP server that empowers AI assistants (Cursor, Claude, C
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx paper-readerConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Paper Reader
The Paper Reader MCP server gives Claude and other AI assistants direct access to arXiv — the preprint repository hosting millions of scientific papers across physics, mathematics, computer science, and more. Through two focused tools, Claude can search arXiv by keyword and category, then retrieve full paper text in paginated Markdown format, making it practical to read, summarize, and compare research papers within a single conversation.
Prerequisites
- Python 3.11 installed (the server targets Python 3.11 specifically)
- pip or a compatible package manager available
- A Claude desktop client that supports MCP (Claude Desktop, Cursor, or Cherry Studio)
- Internet access to reach arxiv.org APIs
Clone the repository
Clone the paper_reader repository and enter the project directory.
git clone https://github.com/itshen/paper_reader.git
cd paper_readerInstall Python dependencies
Install the required packages from requirements.txt using Python 3.11.
pip install -r requirements.txtReview configuration options
The server reads config.yaml at startup. Key settings include the server port (default 8633), storage directory for cached papers, maximum cache size, and paper retention period.
# config.yaml (defaults shown)
server:
name: paper-reader
port: 8633
auth:
default_password: admin123
salt: change-this-salt
storage:
data_dir: ./data
papers:
max_size_mb: 1024
max_age_days: 90Start the MCP server
Launch the server using the provided run script or directly with Python 3.11.
# Using run script
./run.sh
# Or directly
python3.11 server.pyAdd the server to Claude Desktop config
Register the Paper Reader MCP server in claude_desktop_config.json.
{
"mcpServers": {
"paper-reader": {
"command": "python3.11",
"args": ["/absolute/path/to/paper_reader/server.py"]
}
}
}Change the default admin password
The server ships with default credentials (admin / admin123). Log in to the admin panel at http://localhost:8633 and change the password before using the server on any shared or networked machine.
Paper Reader Examples
Client configuration
claude_desktop_config.json snippet for the Paper Reader MCP server.
{
"mcpServers": {
"paper-reader": {
"command": "python3.11",
"args": ["/Users/you/paper_reader/server.py"]
}
}
}Prompts to try
Example research requests you can send to Claude once the server is running.
- "Search arXiv for papers about retrieval-augmented generation published in 2024"
- "Search for transformer attention papers in the cs.CL category"
- "Get the full text of arXiv paper 1706.03762 (Attention Is All You Need)"
- "Summarize the methodology section of arXiv paper 2305.10601"
- "Find the top 5 papers on diffusion models and compare their approaches"Troubleshooting Paper Reader
Server fails to start with a Python version error
The server targets Python 3.11 specifically. Run python3.11 --version to confirm the correct version is installed. If using pyenv or conda, activate the 3.11 environment before running the server.
search_papers returns no results for known papers
arXiv search queries are case-sensitive for category filters. Ensure category values like cs.CL, cs.AI, or quant-ph use the exact capitalisation used by arXiv. Broad keyword searches without categories tend to be more reliable.
get_paper_content returns truncated text for long papers
Use the page parameter to retrieve subsequent pages: get_paper_content('paper-id', page=2). The max_chars parameter (default 50000) controls how much text is returned per page.
Frequently Asked Questions about Paper Reader
What is Paper Reader?
Paper Reader is a Model Context Protocol (MCP) server that mcp server that empowers ai assistants (cursor, claude, cherry studio) to search, read, and summarize arxiv papers directly. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Paper Reader?
Follow the installation instructions on the Paper Reader GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Paper Reader?
Paper Reader works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Paper Reader free to use?
Yes, Paper Reader is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
Paper Reader Alternatives — Similar Search & Data Extraction Servers
Looking for alternatives to Paper Reader? Here are other popular search & data extraction servers you can use with Claude, Cursor, and VS Code.
TrendRadar
★ 58.0kA real-time hotspot monitoring and news aggregation assistant that provides AI-powered analysis of trending topics across multiple platforms via the Model Context Protocol. It enables users to track news and receive automated notifications through va
Scrapling
★ 52.7k🕷️ An adaptive Web Scraping framework that handles everything from a single request to a full-scale crawl!
PDF Math Translate
★ 33.9k[EMNLP 2025 Demo] PDF scientific paper translation with preserved formats - 基于 AI 完整保留排版的 PDF 文档全文双语翻译，支持 Google/DeepL/Ollama/OpenAI 等服务，提供 CLI/GUI/MCP/Docker/Zotero
GPT Researcher
★ 27.2kAn autonomous agent that conducts deep research on any data using any LLM providers
Agent Reach
★ 20.1kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
Xiaohongshu
★ 13.7kMCP for xiaohongshu.com
Browse More Search & Data Extraction MCP Servers
Explore all search & data extraction servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Paper Reader in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Paper Reader?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.