Papers to Markdown - Academic PDF Converter

Convert academic papers from PDF to Markdown or EPUB format with AI-powered OCR and LaTeX support.

🚀 Features

• AI-Powered PDF Conversion: Transform research papers and academic PDFs into clean Markdown or EPUB formats
• LaTeX Equation Support: Preserve mathematical equations with multiple rendering options (MathML, SVG, or clipping)
• Smart Table Extraction: Convert complex tables to HTML or image clipping
• Multi-Language OCR: Support for English and Chinese documents
• Configurable OCR Models: Choose from tiny to gundam-sized models based on your accuracy needs
• Batch Processing: Automatically process multiple PDFs in your directory
• GPU Acceleration: Leverages PyTorch and CUDA for fast processing

📋 Requirements

• Python 3.13 or higher
• CUDA-compatible GPU (recommended for faster processing)
• UV package manager

🔧 Installation

1. Clone this repository:

git clone https://github.com/ahnafnafee/papers-to-markdown.git
cd papers-to-markdown

2. Install dependencies using UV:

uv sync

3. Activate the virtual environment:

# Windows
.venv\Scripts\activate

# Linux/Mac
source .venv/bin/activate

💻 Usage

Basic Markdown Conversion

Convert PDFs to Markdown format:

uv run main.py --format markdown --output-dir markdown

EPUB Conversion

Convert PDFs to EPUB format:

uv run main.py --format epub --output-dir epubs

Advanced Options

uv run main.py \
  --root . \
  --format markdown \
  --output-dir output \
  --ocr-size base \
  --lang en \
  --latex-render svg \
  --table-render html

⚙️ Configuration Options

Option	Values	Default	Description
--root	Path	.	Directory containing PDF files
--output-dir	Path	epubs	Output directory for converted files
--format	epub, markdown	epub	Output format
--ocr-size	tiny, small, base, large, gundam	small	OCR model size (larger = more accurate)
--lang	en, zh	en	Document language
--latex-render	mathml, svg, clipping	svg	LaTeX equation rendering mode (EPUB only)
--table-render	html, clipping	html	Table rendering mode (EPUB only)
--models-cache-path	Path	.pdf_craft_models	Model cache directory
--analysis-root	Path	.pdf_craft_work	Work directory for analysis artifacts
--dry-run	Flag	False	Preview files without converting

📁 Project Structure

papers-to-markdown/
├── main.py                 # Main conversion script
├── pyproject.toml         # Project dependencies (UV)
├── README.md              # This file
├── .gitignore            # Git ignore rules
├── .venv/                # Virtual environment
├── .cache/               # HuggingFace and matplotlib cache
├── .pdf_craft_models/    # Cached OCR models
├── .pdf_craft_work/      # Temporary analysis files
└── markdown/             # Output directory for Markdown files

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

This project uses pdf-craft for PDF processing.

🔍 Keywords

academic paper converter, PDF to Markdown, PDF to EPUB, research paper OCR, LaTeX equation preservation, academic document processing, AI PDF converter, scientific paper conversion, Python PDF tool, batch PDF conversion

📞 Support

For issues and questions, please open an issue on GitHub.

---

Made with ❤️ for researchers and academics