Research Papers MCP

Seamlessly search and analyze academic papers from arXiv through the Model Context Protocol

2M+ Papers Available

Real-time Search Results

MCP Protocol Support

Example Request

# Search for papers on quantum computing
POST /mcp
Content-Type: application/json

{
  "method": "tools/call",
  "params": {
    "name": "search_papers",
    "arguments": {
      "topic": "quantum computing",
      "max_results": 5
    }
  }
}

Quick Start

Get up and running with Research Papers MCP in minutes

claude_desktop_config.json

{
  "mcpServers": {
    "research-papers": {
      "command": "npx",
      "args": ["-y", "@saidiibrahim/research-papers-mcp"],
      "env": {}
    }
  }
}

search_papers.py

import httpx
import json

# Initialize connection
client = httpx.Client(base_url="http://localhost:8787")

# Search for papers
response = client.post("/mcp", json={
    "jsonrpc": "2.0",
    "method": "tools/call",
    "params": {
        "name": "search_papers",
        "arguments": {
            "topic": "machine learning",
            "max_results": 5
        }
    },
    "id": 1
})

papers = response.json()
print(json.dumps(papers, indent=2))

search_papers.ts

import { Client } from '@modelcontextprotocol/sdk';

// Initialize MCP client
const client = new Client({
  name: 'research-papers-client',
  version: '1.0.0',
});

// Connect to server
await client.connect({
  transport: {
    type: 'sse',
    url: 'http://localhost:8787/sse'
  }
});

// Search for papers
const result = await client.callTool({
  name: 'search_papers',
  arguments: {
    topic: 'quantum computing',
    max_results: 5
  }
});

console.log(result);

terminal

# Search for papers using cURL
curl -X POST http://localhost:8787/mcp \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "params": {
      "name": "search_papers",
      "arguments": {
        "topic": "artificial intelligence",
        "max_results": 5
      }
    },
    "id": 1
  }'

API Endpoints

/sse

Server-Sent Events connection for real-time MCP streaming

/mcp

Streamable HTTP connection for MCP protocol

Available Tools

search_papers

Search for academic papers on arXiv and get complete information including abstracts, authors, and download links.

• topic (string) - The topic to search for (e.g., 'machine learning', 'quantum computing')

• max_results (number) - Maximum number of results to retrieve (1-20, default: 5)

Prompt Templates

generate_search_prompt

Generate a comprehensive prompt for searching and analyzing research papers on a specific topic.

• topic (string) - The research topic to search for

• num_papers (string) - Number of papers to retrieve (1-20, default: 5)

How to Connect

Configure Your Client

Add this server to your MCP-compatible client configuration

Choose an Endpoint

Connect via SSE for streaming or HTTP for standard requests

Start Searching

Use the tools to search and analyze academic papers instantly