How to Convert PDF to EPUB: Step by Step

PDF-to-EPUB conversion is a lossy transformation by design: PDF has fixed pages, EPUB doesn't. The converter extracts the text and images from the PDF and re-typesets them into reflowable HTML. What you get is easier to read on any screen — but it won't look identical to the source. Here's how to do it and what to expect.

When This Actually Makes Sense

Convert a PDF to EPUB when you want to read a text-heavy document on a phone, e-reader, or tablet. PDFs render poorly on small screens — fixed page size means constant zooming and scrolling. An EPUB of the same content adjusts to any width and lets you change font size.

Don't bother converting: forms, brochures, sheet music, academic papers with precise diagrams, anything where layout carries meaning. Keep those as PDFs. See the EPUB vs PDF comparison if you're not sure which applies.

Before You Start: Is Your PDF Actually Text?

Open the PDF and try to select a paragraph. If you can't select text, it's a scan — images of pages rather than text. Converters can't extract text from scanned PDFs. You'll need to run OCR first (Adobe Acrobat, ABBYY FineReader, or free tools like Tesseract/OCRmyPDF), then convert the OCR'd PDF.

Method 1: Epublys (Browser, No Install)

1. Open the PDF to EPUB converter
2. Upload your PDF — up to 20MB free, 100MB with Pro
3. Click Convert to EPUB
4. Download — valid EPUB 3, automatic chapter detection, no watermarks, in-memory processing only

After converting: edit the metadata to set the correct title and author (PDFs often have generic or missing metadata), then validate before loading onto your e-reader.

Method 2: Calibre (More Control)

Use Calibre when the automatic chapter detection misses breaks, or when you need to adjust fonts, margins, or CSS output. Free desktop app, 200MB install.

1. Add your PDF via "Add books"
2. Select it and click Convert books
3. Set output format to EPUB
4. Under "Structure detection": set the chapter detection XPath if headings aren't being recognized
5. Under "Look & feel": set font size, margins, and line spacing
6. Click OK — export the EPUB from the book's context menu

Calibre is also the right tool if the Epublys output has layout issues you want to fix manually — open the EPUB in Calibre's editor after converting.

Method 3: Command Line

pandoc input.pdf -o output.epub --metadata title="Book Title"

Pandoc handles text-heavy PDFs well. Chapter detection is basic — it works best when the PDF has consistent heading markup. For automation pipelines, the Epublys REST API accepts base64-encoded PDFs and returns EPUBs.

What Converts Well vs. What Doesn't

• Transfers cleanly: prose text, basic headings, inline images, simple tables, hyperlinks
• Gets simplified: multi-column layouts (become single-column), precise CSS positioning (dropped), floated sidebars (moved inline)
• Fails entirely: scanned pages without OCR, password-protected PDFs, PDFs where the "text" is actually paths (some generated files)

Fixing Common Problems

Output is too large: Image-heavy PDFs produce large EPUBs. Compress the EPUB after converting — image optimization typically cuts 30–60% of the file size.

Chapter breaks are wrong: Adjust heading detection in Calibre's structure detection settings, or use the Epublys converter options to set a fixed page-count per chapter.

File is too big to convert free: Most text PDFs are well under 20MB. If yours is larger, it's likely image-heavy — consider stripping non-essential images from the PDF first, or upgrade to Pro for the 100MB limit.

After Converting

• Set the metadata — title, author, description, cover image
• Validate — catch any structural errors before loading to a device
• Send to Kindle: email it to your Kindle address
• Compress if the file is larger than you need

Related Guides

• EPUB vs PDF — when conversion makes sense
• EPUB to PDF — the reverse direction
• How to send EPUB to Kindle
• What is an EPUB file?
• All free EPUB tools