iloveepubProcessed locally — files never leave your device
EPUB to KEpub
Optimize any EPUB for Kobo e-readers by converting it to KEpub, Kobo's enhanced format with chapter-level reading statistics and in-chapter progress. The conversion runs entirely in your browser — your book never leaves your device. Free, no accounts, no limits.
01Select your EPUB
02Convert & download
When a Kobo is treating your sideload as a plain EPUB
Kobo store purchases are already KEpub, and they are usually DRM, which this page will not open. The books that need the conversion are the ones you copied over USB: a Project Gutenberg novel, a magazine export, a book you compressed on this site. Until the file carries Kobo's sentence spans, a Clara, Libra, Sage, or Elipsa shows ordinary page counts and skips the per-chapter time-remaining line.
The filename is part of the job. Kobo turns the enhanced renderer on for a file that ends in .kepub.epub. A file named .epub can contain every span and still be opened as a plain EPUB. The download uses the double extension for that reason.
The markup this conversion actually inserts
Each sentence becomes a span with class koboSpan and an id such as kobo.4.2. The paragraph number and the sentence number are how the device counts time left in the chapter. CJK punctuation splits the sentence immediately. An ASCII period, question mark, or exclamation mark splits only when whitespace, the end of the text, an opening quote, or a CJK character follows, so strings like 3.14 and e.g. stay intact. The words are not rewritten. The spans are wrapped around the same text.
script, style, pre, audio, video, svg, and math are left untouched. The body is wrapped in div#book-columns and div#book-inner, and one stylesheet line is added so Kobo's default body margin does not stack. In the OPF, the cover item gains the cover-image property if it lacked it, and meta tags whose name starts with calibre: are removed. A chapter that is already a KEpub is skipped, so spans are not nested. A chapter that is not well-formed XML is copied unchanged and listed as skipped. The cap is 120 MB.
What you still will not get
This is not a Kobo driver. It does not write series information into the device database, it does not manage collections, and the statistics appear only after the Kobo itself has counted the spans. There is no conversion back from KEpub to a plain EPUB. Fixed-layout and picture books gain little, because there are few sentences to mark.
A skipped chapter stays in the book, without sentence-level stats. DRM files from the Kobo store are refused. Removing that DRM is not something this page does.
Kepubify and Calibre's Kobo driver
The standalone kepubify tool and Calibre's KoboTouchExtended plugin perform the same class of markup transform, and Calibre can also copy the file over USB and update the device library. If you already manage the shelf in Calibre, the driver is the shorter path.
This page does only the markup. It does not talk to the Kobo, and it removes calibre: metadata while patching the OPF. One file, no install: that is the whole job it is for.
How to convert EPUB to KEpub
Select your EPUB file
Open the book from your device. It is read by your browser directly — nothing is uploaded anywhere.
Convert to KEpub
Every sentence gets its Kobo marker and the book is repacked as a .kepub.epub file. A progress bar shows each step as it runs locally.
Download and copy to your Kobo
Save the converted book, then sideload it to your Kobo over USB. The device recognizes the enhanced format automatically.
Frequently asked questions
Why does the download end in .kepub.epub?
Kobo chooses the enhanced renderer from that double extension. A file named .epub is often opened as a plain EPUB even if the spans are present, and the chapter statistics stay off.
What is inserted into a chapter?
Each sentence becomes a span with class koboSpan and an id like kobo.4.2. The body is wrapped in div#book-columns and div#book-inner, and a one-line stylesheet stops Kobo's default margin from stacking. The words themselves are not rewritten.
What if the book is already a KEpub?
A document that already has book-columns or koboSpan is left alone, so spans are not nested. If every document is already converted, the tool stops and says so instead of writing a second copy.
Do I still need Calibre?
Only if you want Calibre to send the book over USB and keep series information in the Kobo database. The span transform itself does not require Calibre. kepubify and the KoboTouchExtended plugin do equivalent markup work on the desktop.
A chapter was skipped. Is the download missing pages?
No. That chapter failed XML parsing or had no body, so it was copied unchanged and will not have sentence-level stats. The other chapters still have their spans. The result line reports how many files were skipped.