Reflow

Make a PDF readable on your e-reader.

Other converters keep the page layout, so the text lands on top of itself. Reflow rebuilds the paragraphs, chapters and figures instead.

Reflow - Convert, split, and edit PDFs into clean EPUBs privately. | Product Hunt
Drop a PDF here or tap to choose one · nothing leaves this browser

How it works

  1. Drop in the PDFThe whole conversion runs on your device. There's no upload, no account and no size limit beyond your own memory.
  2. Check the previewYou read one chapter at a time, set the way your e-reader will set it. Reflow takes chapter names from the book's own contents page or bookmarks.
  3. Fix the contentsRename, reorder, merge or remove entries in the list. Click any line in the preview to promote it to a chapter heading, or click a heading to demote it.
  4. Download and side-loadYou get a standard EPUB 3 with a working table of contents, a cover and your images.

Questions

Why does my PDF look broken on an e-reader?

A PDF fixes every word to a coordinate on a fixed-size page. An e-reader needs to reflow text to fit its screen and your chosen font size, and it can't, so you end up pinching and scrolling across a page built for A4. Converters that keep those coordinates produce EPUBs where lines sit on top of each other. Reflow uses the coordinates only to work out reading order, then discards them and rebuilds real paragraphs.

Is my file uploaded anywhere?

No. There is no server to upload to. The PDF is read by your own browser, and the EPUB is assembled on your device. You can disconnect from the internet after the page loads and it still works.

My PDF is a scan and nothing happens. Why?

A scanned PDF contains photographs of pages, not text, so there is nothing to extract. Reflow tells you when it finds no text layer. You need OCR software to read the pages first; free options include running the file through OCRmyPDF, or opening it in Google Drive, which runs OCR when it converts to a Google Doc.

The chapters came out wrong. Can I fix them?

Yes, and you don't have to re-convert. Reflow takes chapter names from the PDF's bookmarks, or from the book's own contents page if it has one. When neither exists it guesses from the typesetting, and typesetting varies enormously. In the contents panel you can rename, reorder, merge and remove entries, and clicking any line in the preview promotes it to a chapter heading.

Do images survive?

Yes. Figures are extracted and placed back at the point in the text where they appeared on the page, rather than collected at the end or dropped. A large image on an early page is used as the cover. Artwork that repeats on most pages is treated as a logo or watermark and left out.

How do I get the EPUB onto a Kindle or Kobo?

On a Kobo, plug it into a computer and copy the file into the device's folder; it appears in your library. On a Kindle, email the file to your Send to Kindle address, or use the Send to Kindle page in a browser — modern Kindles accept EPUB directly. Calibre handles both if you prefer a desktop tool.

What won't it handle?

Multi-column pages interleave into nonsense, because lines are grouped by their vertical position. Tables, footnotes and hand-set layout lose their structure. Very large books are bounded by your device's memory. It is built for books and documents that are mostly prose.

Is it free? Is there a catch?

It's free, there's no account, no limit on file size beyond your own memory, and no watermark. There is nothing to monetise because nothing reaches a server. The source is on GitHub.

Text and images only — hand-set layout, tables and multi-column pages won't survive intact. A scanned PDF has no text layer at all and needs OCR before any converter can read it.