How to download novels with WebToEpub
By the Novel Translator team. Last updated 15 November 2026.
WebToEpub is a free, open-source browser extension for Firefox and Chrome, made by David Teviotdale (licence GPLv3). It reads the chapter pages of a web novel and packs them into one EPUB file on your computer. Novel Translator did not make WebToEpub and is not connected to it. We checked the steps below against the project's README and FAQ on 15 November 2026.
Before you start: only copy what you may keep
An EPUB made with WebToEpub is a copy of the site's text. Many sites do not allow copies in their terms, and paid or locked chapters belong to the author and the publisher. Use WebToEpub for texts in the public domain, for your own writing, or for stories whose author or site allows offline copies. Check the terms of each site first. We do not list or link to sites that post translations without a licence.
For a public-domain work, look for the library's own download first. Each work page on Wikisource (for example 鲁迅《狂人日記》 on Chinese Wikisource, or 이상「날개」 on Korean Wikisource) has an EPUB download link, so you do not need an extension. Aozora Bunko, the Japanese public-domain library, offers its works as ZIP files with a TXT file inside.
Install WebToEpub
Install it from the Chrome Web Store or from Firefox Add-ons. Use the store links on the project page, so that you get the real extension and not a copy. Its icon is at the top right of the browser window. In Chrome, you may need to pin it from the extensions menu first.
Make an EPUB in five steps
- Open the story. Go to the page that lists the chapters (the table of contents). On some sites, the first chapter page works too.
- Click the WebToEpub icon. WebToEpub shows the story details and the list of chapters that it found.
- Check the details. Correct the story details, such as the title and the author, if they are wrong. Choose an image for the cover.
- Select the chapters."Select All" selects every chapter. To keep only some chapters, click "Edit Chapter URLs": you get one link per chapter, and you can delete links, change their order or change a chapter title. Then click "Apply Changes".
- Click "Pack EPUB". A progress bar shows while WebToEpub fetches each chapter and its images. When it is full, the EPUB is in your downloads folder.
When WebToEpub does not know the site: the Default Parser
WebToEpub has a parser for each site it knows. For other sites it opens the Default Parser, where you tell it which part of the page holds the story. You do this with a CSS selector: a short name for one element of the page, such as div.post_content.
The Default Parser has five fields:
- Hostname: filled in for you.
- URL of first chapter: used by the Test button. You can leave it empty.
- CSS Selector for the content: the element that holds the chapter text.
- CSS Selector for Title of Chapter: only if the title is outside the content element.
- CSS Selector for Elements to remove: ads, share buttons and other parts you do not want.
To find the selector:
- Open one chapter in the browser.
- Right-click the first word of the chapter text, then click "Inspect".
- In the panel that opens, move up through the parent elements until one element holds all of the chapter text.
- Write its selector: the tag name, a dot and its class, for example
div.main_text. - Put it in the content field and click "Test". WebToEpub fetches the first chapter and shows what the EPUB will contain.
- If the result is correct, click "Finished" and continue with step 4 above.
Worked example: a public-domain page from Aozora Bunko
We looked at the HTML of 芥川龍之介「羅生門」(1915) on Aozora Bunko on 15 November 2026. These are the values you would put in the Default Parser. We checked them against the page's HTML; we did not run WebToEpub for this guide.
- CSS Selector for the content:
div.main_text. The whole story is in this one element. - CSS Selector for Title of Chapter:
h1.title. The title is outside the content element. - CSS Selector for Elements to remove:
rt, rp. The page has 129 ruby readings, such as げにん over 下人. Remove them if you want only the main text.
Why remove the readings before a machine translation: without the ruby layout, a reading can show as plain text next to its word, as in 下人(げにん). The word then appears twice in the source text.
Common problems
- "No parser found": use the Default Parser as shown above.
- Unwanted text in the chapters: add its selector to "Elements to remove". If it has no clear selector, the WebToEpub FAQ suggests removing it from the finished EPUB with EpubEditor, a tool by the same author.
- The site slows you down: some sites limit fast requests. If a site answers with HTTP 429 (too many requests), WebToEpub slows down as the site asks.
- Chapter lists from NovelUpdates: not supported. The FAQ says the site owners asked the author not to add this.
- Scrambled or protected text: WebToEpub does not decrypt it.
Translate the EPUB
If the book is in Chinese, Japanese or Korean and you are allowed to translate it, you can upload the EPUB to Novel Translator. The upload takes EPUB and TXT files of up to 30 MB each.
- If the EPUB is larger than 30 MB, split it with EPUB Split & Merge.
- A TXT file must be UTF-8. Aozora Bunko's TXT files use Shift_JIS, and older Chinese and Korean TXT files often use GBK, Big5 or EUC-KR. Run them through the Novel File Cleaner first. In our test on 15 November 2026, it changed a Shift_JIS, a GB18030, a Big5 and an EUC-KR file into UTF-8 files that the upload accepts.
- Check the cost with the pricing calculator before you start.
What to expect for each language: Chinese web novels, Japanese light novels and Korean web novels.
Sources
- WebToEpub on GitHub (README: installation, usage, licence)
- WebToEpub FAQ (chapter ranges, Default Parser, rate limits, NovelUpdates)
