You have a folder full of PDFs, EPUBs, DOCXs, and ancient RTFs.
You need clean, readable Markdown.
You don't want to think about it.

That's TOMD.

Drop it in a directory. Run tomd. Done.

TOMD is a blazing-fast, terminal-native document converter that turns virtually any document format into clean, LLM-ready Markdown — with a gorgeous live progress display, per-file status tracking, and page-by-page PDF progress reporting.

No GUIs. No configuration files. No nonsense.

The recommended way — install globally with pipx so it's available everywhere on your system:

# Install pipx if you don't have it
brew install pipx && pipx ensurepath

# Clone the repo
git clone https://github.com/phuzzled/tomd.git
cd tomd

# Install globally (editable — code changes take effect immediately)
pipx install -e .

You can now run tomd from any directory on your machine.

Prerequisites: Install Pandoc for most formats. For Kindle files, install Calibre.

brew install pandoc
brew install --cask calibre

# Convert all supported files in the current directory
tomd

# Convert specific files
tomd report.pdf brief.docx chapter.epub

# Recurse into subdirectories
tomd -r

# Force overwrite of existing .md files
tomd --force

# Write output to a specific directory
tomd --output-dir ./converted

# Combine options
tomd -r --force --output-dir ./markdown

# Show version
tomd -v

All converted files land in a md-done/ folder (or wherever you point --output-dir), leaving your originals exactly where they are.

TOMD uses a sleek, live-updating terminal display powered by Rich:

• Live results table — every file ticks ✓ green, ⊘ yellow (skipped), or ✗ red (failed) as it completes
• File progress bar — tracks how many files have been processed
• PDF page bar — for large PDFs, a █░░░░░░░░░ bar shows page-by-page progress so you're never left wondering
• Final summary — a clean table of every result when the run is complete

No blinking cursors. No silent freezes. Just pure, observable progress.

TOMD's PDF engine is powered by PyMuPDF4LLM, specifically designed to produce structured, clean Markdown that preserves:

• Headings, bold, italic, tables
• Reading order across multi-column layouts
• Inline code and technical notation

Feed the output directly into your RAG pipelines, vector databases, or AI workflows. TOMD gets the docs in; your model does the rest.

git clone https://github.com/phuzzled/tomd.git
cd tomd

# Install in editable mode (changes to the source reflect immediately)
pipx install -e .

# Or use a virtual environment
python -m venv .venv && source .venv/bin/activate
pip install -e ".[dev]"

# Run tests
pytest

MIT — free to use, modify, and distribute.