How do I use mcp-pdf-tools?

1. Click on "Install Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@mcp-pdf-tools Summarize the key points in /documents/annual-report.pdf" That's it! The server will respond to your query, and you can continue using it as needed. Here is a step-by-step guide with screenshots.

mcp-pdf-tools

MCP server for extracting text, searching, and analyzing PDF files

npm version License: MIT

Give Claude the ability to read, search, and analyze PDF documents through MCP.

Features

pdf_info - Get metadata (title, author, pages, text length)
pdf_extract_text - Extract all text from a PDF
pdf_extract_pages - Extract text from specific page ranges
pdf_search - Search for text within a PDF with context
pdf_word_stats - Word count and top word frequencies

Quick Start

With Claude Desktop

{ "mcpServers": { "pdf-tools": { "command": "npx", "args": ["-y", "mcp-pdf-tools"] } } }

With Claude Code

claude mcp add pdf-tools npx mcp-pdf-tools

Tools

pdf_info

Get metadata about a PDF.

pdf_extract_text

Extract all text. Supports max_chars to limit output size.

pdf_extract_pages

Extract text from page range (start_page to end_page, 1-indexed).

pdf_search

Search for text with surrounding context. Case-insensitive.

pdf_word_stats

Get word count and top N most frequent words.

Example Conversation

You: Summarize the key points in /documents/report.pdf
Claude: Let me read that PDF... Uses pdf_info to check the document Uses pdf_extract_text to read the content
Here are the key points from the 24-page report...