Open-Source Licences
Last updated: 13 October 2026
MarkPrep is built on open-source software. Each library below is listed with its licence and its job in MarkPrep. We thank the people who maintain them.
The website (runs in your browser)
- pdf.js (pdfjs-dist) (Apache-2.0): reads text from PDF files
- Tesseract.js (Apache-2.0): runs OCR on scanned pages and images
- tesseract.js-core (Apache-2.0): the Tesseract OCR engine, built as WebAssembly
- mammoth.js (BSD-2-Clause): turns Word .docx files into HTML. Copyright (c) 2013, Michael Williamson.
- SheetJS Community Edition (xlsx) (Apache-2.0): reads Excel spreadsheets
- Turndown (MIT): turns HTML into Markdown
- turndown-plugin-gfm (MIT): adds GitHub-Flavored Markdown tables to Turndown
- PapaParse (MIT): reads CSV and TSV files
- JSZip (MIT (dual MIT or GPL-3.0, used under MIT)): opens EPUB and PowerPoint files and builds .zip downloads
- Mozilla Readability (Apache-2.0): keeps the main article when converting a web page
- gpt-tokenizer (MIT): estimates OpenAI token counts with the o200k_base encoding
- marked (MIT): renders Markdown previews
- highlight.js (BSD-3-Clause): colours code blocks in the Markdown viewer. Copyright (c) 2006, Ivan Sagalaev.
- Mermaid (MIT): draws diagrams in the Markdown viewer
- docx (MIT): builds Word .docx files in Markdown to Word
- DOMPurify (MPL-2.0 or Apache-2.0): cleans HTML before it is shown
- Next.js (MIT): the website framework
- React and React DOM (MIT): the user interface library
- Zustand (MIT): keeps the converter's settings and state
- Lucide React (ISC): icons
- clsx (MIT): joins CSS class names
- tailwind-merge (MIT): merges Tailwind CSS classes
- Tailwind CSS (MIT): styling
Data
- Tesseract English language data (tesseract-ocr/tessdata) (Apache-2.0): the English model the OCR reads with
The command-line converter and MCP server (npx markprep)
- Model Context Protocol TypeScript SDK (@modelcontextprotocol/sdk) (MIT): runs the local MCP server
- Commander (MIT): reads the command-line options
- Chokidar (MIT): watches folders for new files
- Zod (MIT): checks the MCP tool's input
- pdf.js (pdfjs-dist) (Apache-2.0): reads text from PDF files
- mammoth.js (BSD-2-Clause): turns Word .docx files into HTML. Copyright (c) 2013, Michael Williamson.
- SheetJS (xlsx) (Apache-2.0): reads Excel spreadsheets
- Turndown (MIT): turns HTML into Markdown
- turndown-plugin-gfm (MIT): adds GitHub-Flavored Markdown tables to Turndown
Each library remains subject to its own licence. Trademarks and copyrights belong to their owners.