Ebooks are made of HTML. We turn them into Markdown.
EPUB is just zipped HTML, which means the content is already structured — you only need a careful converter that respects the spine order and inline formatting. Ours unpacks the archive, walks every chapter in reading order, and converts each one through Turndown with the GFM plugin, so tables stay as tables, lists keep their hierarchy, and inline <code> doesn't become a fenced block.
This is the cleanest path to feeding a long book into an LLM, building a personal study knowledge base, or migrating a self-published ebook into a Markdown-first publishing pipeline.
What is preserved
- Chapter order taken from the OPF spine (the same order an EPUB reader follows)
- Headings (H1 / H2 / H3) and section structure
- Inline formatting — bold, italic, links, inline code
- GFM tables with proper pipe escaping
- Nested ordered and unordered lists
- Block quotes
- EPUB footnotes (
epub:type="noteref") mapped to CommonMark[^id] - Dublin Core metadata exposed as an optional YAML frontmatter block
The result is a clean Markdown ebook that opens correctly in any Markdown editor and can be re-rendered to PDF, HTML, or another EPUB later.