EPUB to Markdown — Convert Books for AI
Extract each EPUB chapter to Markdown. Ideal for feeding textbooks, self-help books, technical books into ChatGPT/Claude for summaries, explanations, Q&A.
Drop a file here or
.epub (EPUB 2.0 / 3.0, no DRM)
Benefits
Each chapter becomes its own Markdown file — feed individually to AI to manage context-window load.
Table of Contents preserved as heading hierarchy. AI understands book structure.
Download all chapters in a single ZIP. Or single merged MD across all chapters.
How to use
- 1Drop a .epub file into the dropzone.
- 2Tool extracts OPF/HTML from the ZIP, parses each chapter.
- 3View the chapter list + token count per chapter.
- 4Click any chapter to preview, or download the full ZIP.
What is EPUB to Markdown?
EPUB is the standard ebook format (after Kindle MOBI/AZW), really a ZIP of XHTML and CSS files. It's an open standard with broad support (Apple Books, Calibre, Adobe Editions, Kindle Cloud Reader). When AI users want to feed an entire book into ChatGPT/Claude for studying, the EPUB needs to be converted to text first.
This tool extracts the ZIP, parses the OPF (Open Packaging Format) for chapter order, converts each XHTML chapter to Markdown via Turndown. Preserves the book's heading hierarchy (Part > Chapter > Section), navigation links become internal anchors.
Use cases: students feeding EPUB textbooks into Claude for summaries; self-learners (IT/business) feeding technical books for chapter explanations; researchers processing non-fiction to extract key insights.
- ✓EPUB 2.0 and EPUB 3.0 support
- ✓OPF spine parsed for correct chapter order
- ✓Table of Contents → MD heading hierarchy
- ✓Images in book extracted as base64 (toggle to skip)
- ✓Footnotes and endnotes attached at end of chapter
- ✓Export single merged MD or per-chapter ZIP
- ✓Total + per-chapter token count to plan AI feeds
When to use it
University textbook EPUB → MD → feed Claude for chapter summaries before exams.
O'Reilly, Manning EPUBs → MD → ChatGPT explains code examples or hard concepts.
Self-improvement books → MD → AI extracts action items or core principles.
Public domain literature EPUB → MD → AI analyses characters, plot, themes.
Subscription books from Scribd, Kindle Cloud → export EPUB → MD → AI summary before full read.
How it works
EPUB is a ZIP containing container.xml pointing to the OPF file; OPF specifies the spine (read order) and manifest (all files). The tool extracts via jszip, reads the OPF for order, takes XHTML per spine item, renders through Turndown.
TOC from NCX (EPUB 2) or nav.xhtml (EPUB 3) is parsed to assign heading levels. A chapter may contain nested sections — we maintain that hierarchy in the Markdown output (## Chapter 1 → ### Section 1.1 → #### Subsection 1.1.1).
DRM (Digital Rights Management): the tool does not bypass DRM. If your EPUB has DRM (purchased from Kindle/Kobo), use Calibre + plugin to remove it first. Public-domain books (Project Gutenberg, Standard Ebooks) and self-generated EPUBs work freely.
EPUB → Markdown FAQ
Are Kindle (.azw, .mobi) files supported?
Not in the free tier. AZW/MOBI need Calibre to convert to EPUB first. Phase 2 will add AZW3 via server worker.
How is DRM-protected EPUB handled?
Not bypassed. The tool will fail to parse. Use Calibre + DeDRM plugin first, or use DRM-free EPUBs (public domain, Standard Ebooks).
Can the browser handle very large books (500+ pages)?
Browsers handle up to ~5 MB EPUBs comfortably. Larger may slow. Per-chapter token counts help you decide whether to feed the whole book or chapter-by-chapter.
Are cover images and artwork preserved?
Yes — converted to inline base64. Toggle 'Strip images' to drop them if you only need text for AI.
Can I extract just one specific chapter?
Yes. After parsing, the chapter list is shown — click any to preview/download individually.