PDF vs EPUB: What's the Difference?
PDF and EPUB both display books and documents, but one locks the page and one lets the text flow. Here is how each works and which one fits your file.
NewKitApp 10 min de lecture
Nos guides existent en anglais et en indonésien. Les articles ci-dessous s'ouvrent en anglais.
Open a PDF and an EPUB of the same novel side by side and you see two books. On a large monitor they look similar. Make the window narrow, and they stop being similar: the PDF shows a cramped slice of a fixed page, while the EPUB rewraps its paragraphs to the new width and keeps reading comfortably.
That single behavior comes from a deep difference in how the two formats think about a page. Once you understand it, every other difference between them, from file size to accessibility to which device feels right, follows from it.
What each format actually is
A PDF describes appearance. The file tells a viewer exactly where every letter, line and image sits, on pages of a fixed size, in coordinates measured in points. The philosophy comes from print: a PDF is a digital printout, and the goal is that it looks identical everywhere, on any screen, in any country, decades from now. Fonts get embedded so the document never depends on what is installed. Because everything is pinned in place, a PDF can hold anything a printing press can: multi-column magazine layouts, forms with fields, signatures, precise diagrams, annotations, even small programs.
An EPUB describes content. The file is actually a small zip archive of XHTML chapters, a stylesheet, and a manifest describing them. It is a small website in a box. Nothing in it fixes where a paragraph breaks; the reading device lays the text out at display time, in the font and size the reader chose. The philosophy comes from the web: content flows, and the device decides the shape.
This is also why the two formats fail differently. A PDF that opens at all looks right everywhere. An EPUB depends on the reading software for its appearance, and older or unusual readers render the same book slightly differently, the same way old browsers rendered websites differently.
Fixed layout versus flowing text
Take a concrete example. You have a 300-page manual with two columns, callout boxes, and code samples.
As a PDF, the manual is a sequence of 300 fixed pages. Column one ends at the bottom of page 40 and continues at the top of page 41, exactly as printed. On a phone, you zoom and pan, or the viewer fits the whole page to the screen and you read it at 9 points. Page 40 is page 40 for everyone, which makes "see page 40" meaningful.
As an EPUB, the same manual is a set of chapters whose text rewraps. On the phone the two columns usually become one, the code samples scroll sideways if they are wide, and the callout box lands wherever it lands. Comfort on the phone is dramatically better. But the printed page numbers dissolve, because there are no pages until a device makes them, and a table that relied on exact column widths can fall apart.
Neither behavior is a defect. They serve different promises. The PDF promises fidelity: what you sent is what they see. The EPUB promises comfort: what they read adapts to how they read.
A short history explains the split
PDF arrived in 1993 out of the print world, from the same company that built PostScript for imagesetters. Its first job was making sure a document created on one computer printed identically on another. Everything about the format, the fixed coordinates, the embedded fonts, the page as the unit of existence, descends from that job. Later it gained forms, annotations, signatures and digital archival standards, but the page never moved.
EPUB arrived in 2007, when the publishing industry merged several competing e-book formats into one open standard. Its parents were the web and the reflowable e-book readers of the 1990s, and its unit of existence is the chapter, a small XHTML document. EPUB 3, the current version, added support for better typography, foreign language markup, MathML for equations, media overlays for synchronized audio, and fixed-layout pages as an option for children's books and comics that refuse to flow.
That lineage still shows in everyday use. PDF tools behave like printing equipment: precise, conservative, obsessed with fidelity. EPUB tools behave like web browsers: flexible, reflowing, occasionally inconsistent between implementations.
Which device likes which
A laptop handles both formats with equal indifference; it has the screen and the power for anything. The interesting cases are the small screens.
On a phone, EPUB reads the way the phone wants to be read: one column, large type, page turn by thumb tap. A PDF on a phone is workable for skimming and poor for long reading, because you either zoom into a slice of the page and scroll in two axes, or shrink the page until the text is small. Reflow modes exist in some PDF apps and help with simple documents, but they approximate; they cannot know the designer's intent.
On dedicated e-readers with electronic ink, EPUB is the native citizen (Kindle excepted in file terms, as above). Battery life, page turns, and sunlight readability all play to flowing text. Most of these devices render PDFs at all only grudgingly, and a technical PDF on a six-inch e-ink screen is an exercise in patience.
On tablets, the gap narrows. A ten-inch screen fits a PDF page legibly, and PDF reading apps with good annotation support make tablets the nicest place to mark up a fixed document. EPUB remains more comfortable for pure reading, but the tablet is the one device where the choice is genuinely a preference.
On paper, of course, PDF wins by definition: it was born to be printed, and printing an EPUB first requires freezing it into pages, which is a conversion, not a property of the format.
Where each one wins
PDF wins when layout is the message. Contracts where the other party must see the same line numbers. Invoices and receipts that must print identically. Brochures, sheet music, comics, magazines, technical drawings, anything where the designer placed elements on purpose. Forms to fill and sign. Documents that must be archived looking exactly as approved.
EPUB wins when reading is the message. Novels, essays, biographies, anything that is a river of text with occasional images. Long manuals meant to be read on screen rather than printed. Material people will read on phones, where fixed pages punish the eyes. Anything a reader will want at large type sizes at midnight.
A useful shortcut: if the document would survive being retyped in a different font without losing anything important, it belongs in EPUB. If retyping it would destroy it, it belongs in PDF.
The practical differences that follow
Text size and reflow. In a PDF, enlarging the text means zooming the whole page and panning around. In an EPUB, the text itself gets bigger and the pages re-form around it. For anyone with tired eyes or dyslexia, this difference decides the format by itself.
File size. A novel as EPUB commonly weighs a few hundred kilobytes, sometimes less than a single cover image from its PDF equivalent. The PDF embeds fonts and full page layout; the EPUB carries plain text and a handful of images. On a library of a hundred books, this is the difference between tens and hundreds of megabytes.
Fonts. A PDF carries its fonts inside. An EPUB usually names a font family and lets the reading device supply it, or embeds fonts for the body while headings fall back. This is why EPUBs look native on every device and also why a designer loses fine control.
Images. Both formats embed images. In a PDF an image is pinned to its page position. In an EPUB an image floats with the text and scales to the screen, which is great for figures and awkward for charts with small labels.
Navigation. Both support tables of contents. EPUB puts its navigation in a standard machine-readable file, so every reader app can show a chapter list and sync reading position. PDFs have bookmarks and internal links that work well when the producer added them, and no standard way to report "how far through the book" you are, which is why PDF reading positions sync poorly across devices.
Accessibility. Flowing text gives screen readers a linear, sensible path, and reflow supports large-type modes natively. A tagged PDF can be highly accessible too, but tagging is optional and often skipped, whereas an EPUB starts as real text with real structure. In practice, EPUB is the safer bet for readers using assistive technology.
Editing and copying. Both allow text selection and copying where the producer allowed it. For editing, PDF is famously resistant: it is a final-form format, and editing means specialized tools. EPUB chapters are XHTML, so anyone who can edit a web page can edit the book. That cuts both ways; it also makes EPUB content easier to extract and reuse.
Longevity. PDF is one of the most stable document formats in existence, with an ISO standard and decades of viewers. EPUB is also standardized (EPUB 3 is the current version) and built from web technology likely to be readable for a long time. Both outlast most proprietary formats. A tie, with different risks: PDFs can carry executable baggage worth sanitizing, while EPUBs are essentially inert zip files of text.
DRM and the bookstore question
Most EPUBs sold online carry DRM, encryption tied to a specific store's apps and devices. A Kindle book is not an EPUB in the file sense, even though the store accepts EPUB uploads through send-by-email services; Kindle uses its own formats. Kobo, Google Play Books, and Apple Books open EPUB directly, with their own DRM when bought from them.
DRM-free EPUBs, common from independent publishers and public-domain libraries like Project Gutenberg and Standard Ebooks, open in anything, including the reader built into this site.
PDFs carry optional encryption as well, usually a password. A password-protected PDF opens in any viewer once you type the password, which is simpler than store-tied DRM but is a different mechanism with different rules. Neither format's DRM speaks the other's language.
Converting between them
PDF to EPUB works well when the PDF is a river of text. A converter extracts the words, rebuilds paragraphs from line positions, detects headings from font size, and packs the chapters. Novels and reports convert cleanly. Scanned pages carry no extractable text, so a good converter embeds them as page images in place, and a better workflow runs OCR first and then converts. Multi-column academic layouts can interleave, because text flows out of a PDF in the order the producer wrote it, which is not always reading order.
The converter on this site runs entirely in your browser and handles all of this, embedding scanned pages as images so nothing is lost. You can proof-read the result in the site's EPUB reader before moving it to a device.
EPUB to PDF is easy mechanically and philosophically strange: you are freezing a flowing book into fixed pages, choosing a page size, font and margin that the format deliberately never fixed. It is the right move when you need to print a chapter or archive an exact copy, and the wrong move if you expect the comfort of EPUB to survive.
Neither conversion is lossless in the way a copy is. Going PDF to EPUB, you trade layout for flow, on purpose. Going the other way, you trade flow for a snapshot. The direction that improves your reading is the direction to convert.
Reading each one in a browser
Both formats read comfortably in a browser today, with no installation, which matters when you are on a machine that is not yours.
For PDFs, a page-turn reader like the flipbook tool binds the document like a printed book: two-page spread on a wide screen, one page on a phone, drag the corner to turn, keyboard arrows, zoom for small print. You can open a file from disk or pass a link, and share a ready-made reading link with anyone.
For EPUBs, a browser reader like the one at epub reader unpacks the book in the tab and paginates it like a native reading app, with a real table of contents and adjustable type.
Both approaches share one property worth stating: the file is processed on your device. Nothing about your reading is uploaded, which is the same privacy property a desktop reader gives you and the opposite of most upload-based converters.
Choosing in ten seconds
Ask what the file is for.
Someone must see exactly what you see, on paper or in an archive: PDF. Someone must read a long text comfortably on whatever screen they hold: EPUB. It is a form, a contract, a sheet of music, a comic, a blueprint: PDF. It is a book, a thesis you will read on a train, a manual nobody will print: EPUB.
And when the file arrives in the wrong one, convert it in the direction that improves the reading. The formats are not rivals; they are two answers to two different questions, and now you know which question your file is asking.