thinkpdf 1.0.2

PDF to Markdown engine for LLMs. Smart table extraction, OCR, MCP server for AI IDEs.

thinkpdf

Convert PDFs to clean Markdown for LLMs. Includes MCP Server for AI coding assistants.

Features

• PDF to Markdown conversion
• MCP Server for AI assistants (Cursor, Antigravity)
• GUI included
• Batch conversion with parallel workers
• Optional: Docling for better table extraction

Installation

pip install thinkpdf

For GUI:

pip install thinkpdf[gui]

Quick Start

GUI

thinkpdf-gui

CLI

thinkpdf document.pdf                    # Convert single file
thinkpdf document.pdf -o output.md       # Specify output
thinkpdf folder/ --batch                 # Convert all PDFs in folder
thinkpdf folder/ --batch --workers 4     # Parallel conversion

Python API

from thinkpdf import convert

markdown = convert("document.pdf")
print(markdown)

MCP Server Setup

Run this to see the config:

thinkpdf setup

Add to ~/.cursor/mcp.json or ~/.gemini/antigravity/mcp.json:

{
  "mcpServers": {
    "thinkpdf": {
      "command": "python",
      "args": ["-m", "thinkpdf.mcp_server"]
    }
  }
}

Then ask your AI: "Read the PDF at D:\docs\manual.pdf"

MCP Tools

Tool	Description
read_pdf	Convert and return content directly (no file created)
convert_pdf	Convert and save to file
get_document_info	Get PDF metadata

Requirements

• Python 3.10+
• PyMuPDF (included)
• Docling (optional, for best quality)

License

AGPL-3.0