Why the file never leaves this tab

Every other free PDF to EPUB converter works the same way. You choose a file, it travels to a server in another country, software there converts it, and the result travels back. The good ones encrypt the transfer and delete the file after a few hours. You are asked to take that on trust, and for a marketing report it is a reasonable thing to trust.

For a manuscript you have not published, a client deliverable, a medical record or an internal document, it is a strange thing to be asked. This converter reads the PDF with the same file API your browser uses to show you a photo before you post it. The bytes are decoded a few centimetres from your eyes, the EPUB is assembled in memory, and the download button hands you a file that was built on your own machine.

There is no server to trust, because there is no server. That is also why there is no size limit and no queue.

What the converter has to guess

A PDF does not store a document. It stores where each glyph sits on a page. It has no idea what a heading is, where a paragraph ends, or that the words repeating at the top of every page are a running header rather than the first line of the text.

So the conversion is a set of educated guesses, and it is worth knowing which ones. Lines are grouped by their baseline. Paragraphs run on until there is a wide vertical gap, or a short line that ends a sentence. A line set noticeably larger than the body text becomes a heading, and a much larger one becomes a chapter break. A word broken across two lines with a hyphen is put back together. Anything that appears at the top or bottom of most pages is treated as furniture and removed, along with bare page numbers.

On prose these rules are close to invisible. That is why the tool is good on reports, whitepapers, manuscripts and lead magnets, and honest about everything else.

When a PDF will not convert well

A scan is not a document. If the pages are photographs of paper, there is no text layer to read and this tool will tell you so rather than hand you an empty book. Run it through OCR first and come back.

Heavy multi-column layouts, marginal footnotes, pull quotes threaded between columns and complex tables all rely on position to carry meaning, and position is exactly what an EPUB throws away. Images are embedded, but they land at the end of the page they came from, because a PDF never recorded how a picture related to the sentence beside it.

If the original was designed rather than typed, a straight conversion will disappoint you, and no converter on the internet will do better. What you want in that case is to rebuild it.

Reflowable and fixed are not the same kind of file

A PDF is fixed. Every line breaks where it broke when the file was made, which is why reading one on a phone means pinching and dragging. An EPUB is reflowable. The reader chooses the font, the size and the width, and the text rearranges itself to fit.

That is the whole reason to convert, and it is also the reason a converted file will never look like the PDF did. You are trading a designed page for a readable one. If you want both, you have to lay the book out again rather than convert it.