Convert a PDF Into a Reflowable EPUB Ebook

Extracts the text with pdf.js and packages it as a valid EPUB with JSZip, so the words reflow to whatever screen and font size you read on. Text only — and converted without the document being uploaded.

About the PDF to EPUB Converter

A PDF is a fixed layout, which is exactly what makes it painful on a small screen: the page stays the size it was designed for and you pan around it. EPUB is the opposite — the text has no fixed page, so it flows to fill whatever width it is given, at whatever size the reader chooses. Converting between them is the difference between fighting a document on a phone and reading it.

The output is a properly formed EPUB 2 package: an uncompressed mimetype declaration in the right position, a container descriptor, a content manifest listing every part, a reading order, and a navigation file so the table of contents works. Any e-reader, tablet or reading application will open it.

Key Features

  • Genuinely reflowable: The text carries no fixed page dimensions, so your reader controls the typeface, size, margins and line spacing rather than the original layout doing it.
  • A standards-compliant package: The archive is assembled with the mimetype stored uncompressed and first, as the specification requires, which is the detail that determines whether a reader accepts the file at all.
  • Working navigation: A navigation file is generated with an entry for every section, so the contents menu in your reader is populated rather than empty.
  • Readable on anything: Kobo, Kindle applications that accept EPUB, Apple Books, Google Play Books and the open-source readers all handle the format.
  • Dramatically smaller files: Stripping a document down to its text usually produces a file a fraction of the original size, which matters on a device with limited storage.
  • Converted privately: Extraction and packaging both happen in the browser from libraries held on this domain, so a document you are converting for personal reading is not copied to anyone's server.

How to Use the PDF to EPUB Converter

  1. Add the PDF you want to read as an ebook.
  2. Press Convert to EPUB and let the text extraction run through the pages.
  3. Download the .epub file.
  4. Transfer it to your e-reader, or open it in a reading app on your phone or tablet.

How It Works

Two things about the structure will be immediately visible when you open the result, and both are worth knowing before you convert a long document.

The conversion works page by page, and each PDF page becomes its own section with a heading reading Page 1, Page 2 and so on. A three-hundred-page book therefore produces a contents list of three hundred entries named after page numbers rather than chapters. The text reads through continuously, so this affects navigation rather than reading, but it is not the tidy chapter list you might expect. Chapter detection would require inferring structure from typography, which is unreliable enough that the tool does not attempt it.

The metadata is also generic: the title is taken from the filename, the language is recorded as English regardless of the actual content, and the author field is filled with a placeholder rather than the real author. If the book is going into a library where metadata matters, expect to correct it in a tool like Calibre afterwards.

Images are not carried across at all — this extracts text and nothing else — and a scanned PDF will produce an essentially empty ebook, since there is no text in it to extract.

When You Need It

A reader with dyslexia finds standard PDF typography genuinely difficult, and the usual adjustments that help — a wider letter spacing, a specific typeface, a shorter line length, a larger size — are all impossible in a fixed layout. Reflowable text hands those controls to the reader. Since the documents in question are often work material or personal correspondence, converting without an upload keeps a private accessibility need private, which is not a small thing.

An employee wants to read a long internal strategy document properly on a flight rather than skimming it at a desk, and their e-reader takes EPUB. The document is confidential to the company, so pushing it through a free online converter to get it onto a personal device would be a straightforward policy breach — one that a local conversion avoids entirely.

FAQ

Frequently Asked Questions

Why is every chapter in the contents called Page 1, Page 2 and so on?
Because the conversion splits by PDF page rather than by chapter. A PDF does not record where chapters begin in a way that can be read reliably, so the tool uses the one structural boundary it can trust. The text still reads straight through; only the navigation list is affected. Recompiling the EPUB in Calibre lets you define real chapters if it matters.
The author name in my e-reader is wrong.
The metadata written into the file is generic — the title comes from the filename and the author field holds a placeholder rather than the actual author, because a PDF often does not record one accurately. Any ebook management tool will let you edit those fields, and it takes a moment.
My converted EPUB is empty or nearly empty.
The PDF has no text layer, which means it is a scan — images of pages rather than characters. There is nothing for the extraction to find. Confirm it by running the file through the text extractor: if that comes back blank too, the document would need character recognition before any ebook conversion could work.
Are the images from the PDF included?
No. This converts text only, so figures, photographs and diagrams are absent from the ebook. For an illustrated document that dependency on visuals makes EPUB a poor destination anyway — reading the original PDF, or extracting the images separately, will serve you better.
Does converting a confidential document here expose it?
No. pdf.js reads the text in this tab and JSZip builds the ebook archive in browser memory, with both libraries served from this site rather than fetched externally. An internal report or a personal document is converted on your own machine, which is the whole reason to do it here rather than through a service that requires an upload first.
Will the formatting from the PDF carry over?
Only minimally. Paragraph breaks survive, but bold, italics, headings, tables and columns do not — the extraction produces plain text, which is then marked up as simple paragraphs. This is a readability conversion rather than a faithful one, which is the correct trade when the goal is reflowable text.
Keep going

Related Tools

PDF Text ExtractorEPUB to PDF ConverterPDF Metadata Editor