PDF to EPUB

If your PDF is a scan, this will produce an empty book

Check this first, because it is the commonest way the job fails and no converter warns you about it.

There are two kinds of PDF that look identical on screen. One contains real text — you can select a word with your cursor and copy it. The other is a picture of text, produced by a scanner or a phone camera, where the page is one big image and there is no text in the file at all.

This tool reads the text layer. A scanned PDF has none, so you will get an EPUB with chapter headings and nothing underneath them.

The one-second test: open the PDF and try to select a sentence with your mouse. If the text highlights word by word, you are fine. If you get a blue box over the whole page, or nothing at all, it is a scan.

For a scan you need optical character recognition first — Image to Text (OCR) reads text out of an image, and it will at least tell you whether the text is recoverable. Bear in mind that OCR of a whole book is a different and much larger job than converting one that already has text.

What you get, and what the paragraphs will look like

Being straight about this, because it decides whether the result is pleasant to read.

The good part is real: the text genuinely reflows. A PDF has a fixed page size, which is why reading one on a phone means pinching and panning around an A4 sheet. An EPUB has no pages until your reader makes them, so the text sizes itself to your screen and your chosen font. That difference is the entire reason to convert.

The limitation is structural. A PDF does not record where paragraphs begin — it records where every printed line sits on the page. This converter keeps those lines, and each one becomes a separate paragraph in the ebook. So a single paragraph of prose that filled eight printed lines arrives as eight short indented paragraphs.

The words are all there, in order, and it reflows — but it reads loose, with an indent every line or two rather than at each real paragraph break.

Which means the result is good for: reading a document you just need to get through comfortably on a phone or e-reader · a report, contract, manuscript or paper you want at a readable text size · anything where you were previously zooming into a PDF.

And it is not good for: a book you want to keep and reread · anything you intend to publish or sell · a PDF whose lines are already short, such as poetry or a slide deck, where the extra paragraph breaks compound.

If you want a properly structured ebook rather than a readable one, convert here and then tidy the paragraphs in a free ebook editor such as Calibre or Sigil, which is far less work than retyping.

The three chapter modes, and when to abandon the first one

Auto is the default. It starts a new chapter wherever it finds a large vertical gap on the page, on the reasonable assumption that a gap means a heading.

When it works, you get a real table of contents you can jump around with. When it does not, it fails in a visible way: the chapter title is simply whatever line came after the gap, truncated. So a running header, a page number, a figure caption or half a sentence can end up as a chapter name, and a generously spaced document can end up with dozens of chapters.

So after converting, open the table of contents and look at it. Ten seconds, and it tells you whether to keep this version.

One chapter per page is the predictable fallback. Every PDF page becomes a chapter, which is meaningless as chapters but gives you an exact navigation list — genuinely useful for a reference document, a manual or a scanned-and-OCR'd book where you want to jump to page 84.

Single chapter puts everything in one continuous flow with no table of contents. Use it for something short, or when the other two modes produce a mess. It is the mode least likely to surprise you.

Three more things that will not come through

Images. Text only. Diagrams, photographs, charts and the cover are not carried over, so an illustrated book loses the illustrations.

Two-column layouts. An academic paper, a newspaper page or a two-column report will interleave — a line from the left column, then a line from the right — because the text is read in the order the file stores it, not the order you would read it. If your PDF is in columns, the output will be scrambled and no setting fixes it.

Author and language. The book's title comes from the filename, so name the file sensibly before converting. There is no author field, so your reader will show "Unknown Author", and the language is always tagged as English. All three are a thirty-second fix in Calibre if the metadata matters to you.

A password-protected PDF will not open at all — remove the password first.

The EPUB itself is built properly

Worth a mention because this is where quick converters cut corners and you find out when your reader refuses the file.

The output is a standards-compliant EPUB 2.0: the archive's type marker is written first and uncompressed exactly as the specification demands, the package file declares a real manifest and reading order, and the navigation file lists every chapter in sequence. The result opens in Apple Books, Kobo, Google Play Books, Calibre and the rest without complaint.

Nothing is uploaded — the PDF is read on your own device. There is no size limit and no daily quota, because there is no server involved. A very long PDF will take a little while and will use memory while it works, so close other heavy tabs for a 500-page document.

Related tools

EPUB to PDF goes the other way, and its page argues that you usually should not — a fixed-layout PDF is worse on a phone than the EPUB was, which is the same point from the other side. PDF to TXT Converter gives you the plain text with no ebook packaging, which is what you want if the text is going into another document rather than onto a reader.