MCP Server Setup Guide

Connect AI assistants like Claude, ChatGPT, and Cursor to your PDFs — all locally, all private.

What is MCP? MCP (Model Context Protocol) is a standard that lets AI assistants read and work with files on your computer. OpenReader's MCP server gives AI tools access to your PDFs — search, compare, extract text — without uploading anything to the cloud.

Prerequisites

Before setting up the MCP server, make sure you have:

Python 3.10 or newer installed on your computer (download here)
An MCP-compatible AI assistant (Claude Desktop, Claude Code, Cursor, or any tool that supports MCP)

That’s it. No repo to clone, no source code to download — just Python and a one-line pip install.

Step 1: Install the package

Open a terminal and run this one command:

pip install openreader-mcp

That’s it. This installs everything you need — PyMuPDF for reading PDFs, the MCP SDK, and all 14 tools.

Step 2: Test the server

Make sure everything works by running the MCP server directly:

python -m openreader_mcp

If you see no errors, the server started successfully. Press Ctrl+C to stop it — your AI tool will start the server automatically when needed.

Step 4: Configure your AI assistant

Claude Code

Add this to your project's .claude/settings.local.json file (create it if it doesn't exist):

{
  "mcpServers": {
    "openreader": {
      "command": "python",
      "args": ["-m", "openreader_mcp"]
    }
  }
}

Or add it to your global ~/.claude/settings.json to make it available in every project.

Claude Desktop

Open Claude Desktop settings and find the MCP Servers section. Add a new server with:

{
  "mcpServers": {
    "openreader": {
      "command": "python",
      "args": ["-m", "openreader_mcp"]
    }
  }
}

Cursor

Go to Cursor Settings → MCP Servers → Add New Server and enter:

Name: openreader
Type: command
Command: python -m openreader_mcp

What you can ask your AI

Once connected, you can ask your AI assistant things like:

"Open my contract PDF and find the termination clause"
"Compare these two PDFs and tell me what changed"
"Search all my research papers for mentions of machine learning"
"Extract pages 3 through 7 into a separate PDF"
"Merge these three PDFs into one document"
"Compress this PDF to make it smaller"

All available tools

OpenReader's MCP server provides 14 tools for working with PDFs:

extract_text

Read all text from every page of a PDF

get_page_text

Read text from one specific page (1-indexed)

get_metadata

Get PDF info — page count, author, title, file size

get_page_count

Quickly check how many pages a PDF has

search_pdf

Find a word or phrase inside a single PDF with context snippets

compare_pdfs

Show word-level and page-level differences between two PDFs

merge_pdfs

Combine multiple PDF files into one document

split_pdf

Split a PDF into separate single-page files

extract_pages

Pull out specific pages using range notation (e.g. "1-3,5,7-9")

compress_pdf

Create a smaller version of a PDF

index_folder

Build a full-text search index of all PDFs in a folder

search_library

Search across all indexed PDFs at once (fast, keyword-based)

search_semantic

Meaning-based search across indexed PDFs (finds related content)

list_indexed_docs

Show all PDFs currently in the search index

Troubleshooting

"python is not recognized"

Python is not installed or not in your system PATH. Download Python and make sure to check "Add Python to PATH" during installation. Restart your terminal after installing.

"ModuleNotFoundError: No module named 'openreader_mcp'"

The package isn't installed. Run pip install openreader-mcp to install it.

My AI tool says it can't find the MCP server

Make sure python and openreader-mcp are both installed. Try running python -m openreader_mcp manually first to confirm it works. If you're using a virtual environment, make sure your AI tool also uses it.

I installed OpenReader from the Microsoft Store

That's fine! The MCP server is a separate package — just run pip install openreader-mcp and you're good to go. It works independently of the Store app and can read the same PDFs on your computer.