EPUB Translator
Translate unencrypted EPUB books locally, preserving XHTML tags, chapter files, styles and container structure. Text is sent to DeepSeek in short parts; sign-in and free daily translation limits apply.
Examples
Examples use English XHTML snippets and Spanish translations. AI outputs are illustrative only and are not guaranteed. Tags and attributes are shown so you can compare structure before and after.
English
<p>Hello, world!</p>
Translation
<p>Hola, mundo!</p>
Text node inside a paragraph translated.
English
<h1>Chapter One</h1>
Translation
<h1>Capítulo uno</h1>
Heading text translated.
English
<p>Good morning.</p>
Translation
<p>Buenos días.</p>
Short sentence in a paragraph.
English
<em>Thank you</em>
Translation
<em>Gracias</em>
Inline emphasis tag preserved.
English
<a href="note.html">See note</a>
Translation
<a href="note.html">Ver nota</a>
Link text translated, href unchanged.
English
<title>My Book</title>
Translation
<title>Mi libro</title>
XHTML document title text translated.
Understand this tool
EPUB container basics
An EPUB is a ZIP-based container with a package document, spine, manifest, and XHTML content files. Text usually lives inside XHTML elements, so a translator must separate markup from translatable text nodes to avoid breaking chapter order, links, or reading structure.
Why text nodes matter
In XHTML, tags define structure and attributes carry metadata or links. Only text between tags is natural-language content. Translating tag names or attribute values can corrupt links and styles, so a safe EPUB workflow leaves markup intact and translates text nodes in short parts.
Reference guide
| Input | Rule | Result |
|---|---|---|
| <p>Hello</p> | Translate text node; keep p tags | <p>Hola</p> |
| <a href="note.html">See</a> | Keep href; translate link text | <a href="note.html">Ver</a> |
| <em class="x">Yes</em> | Keep class attribute; translate text | <em class="x">Sí</em> |
Tag and attribute handling
The translator treats XHTML as markup: text nodes can be translated, while element names, attribute names, and most attribute values are preserved. Tags such as p, h1, em, and a still wrap the translated text. This helps keep chapter structure and links intact, though exact visual layout is not guaranteed.
File and request limits
Unencrypted EPUB files are limited to 5MB, and expanded archives to 30MB. Extracted text is sent to DeepSeek in parts of no more than 500 characters. The page shows the request count before starting. AI tools require sign-in; each successful request uses one daily free translation or 1 credit. Stop and resume work within the session.
| Input | Rule | Result |
|---|---|---|
| EPUB file | Max size 5MB | Accepted if unencrypted |
| Expanded archive | Max 30MB | Parsed locally |
| Extracted text | Max 20000 chars; sent in <=500-character parts | Request count shown before start |
Ways to use this tool
Translate a public-domain chapter for language study
Upload a small unencrypted EPUB of a public-domain book, review the request count, then translate one chapter at a time. Read the Spanish output beside the original XHTML. Use stop and resume for longer chapters. The tool preserves tags, but layout may differ from a professionally typeset translation.
Prepare a draft for a bilingual reading copy
Convert a short EPUB chapter into Spanish text while keeping paragraphs, headings, and links. Copy each XHTML result back into your own reading copy. Check names, notes, and formatting manually. This is useful for personal drafts, not for promises of publication-ready typesetting or exact visual layout.
Check how inline tags survive translation
Use the example button to load simple XHTML snippets, then translate them. Compare before and after to see whether em, a, and h1 tags remain around the Spanish text. This helps you judge whether a short chapter is safe to process before uploading the full EPUB.
Translate small EPUB notes or documentation
For short internal EPUB files under 5MB, extract text in the browser and send only short parts to the AI. Preserve chapter files and container structure. Review technical terms manually, since automatic outputs may be imperfect and no visual layout guarantee is made. Always verify.
What you can do
Explore the EPUB Translator features and supported options.
Preserve XHTML markup
Translate text inside XHTML chapter files while leaving tags, element names, and most attributes in place. This supports paragraphs, headings, emphasis, links, styles, and the EPUB container structure without promising exact visual layout.
Local EPUB parsing
The EPUB file is parsed in your browser before extracted text is sent for translation. Unencrypted files up to 5MB and expanded archives up to 30MB are accepted. No DRM bypass is provided.
Request count and limits
AI translation requires sign-in. The page shows request count before start, text is sent in parts of at most 500 characters, and each successful request uses one daily free translation or 1 credit.
Stop, resume, copy, download
Stop and resume translation within the session. Use copy, download, example, and reset controls to manage results. Output is text-oriented; exact layout or perfection is not promised.
Frequently asked questions
Related tools
Subtitle Translator
Accept SRT/VTT. Preserve timecodes, cue identifiers and markup. Translate cue text; offer original-format and bilingual downloads.
Image TranslatorAccept PNG/JPEG/WebP, run browser OCR for selected source language, review extracted text then translate; output text only, not edited images.
PDF TranslatorExtract selectable text from PDF locally, preserve page grouping in text output. Export translated TXT, not original-layout PDF. Scanned PDFs require OCR elsewhere.
Word TranslatorAccept DOCX; translate paragraph text while preserving archive structure, tables and images. Complex run styling can consolidate into the first text run.
How to use this tool
Step 1: Upload or paste EPUB text
Choose an unencrypted EPUB file up to 5MB, or paste a short XHTML snippet. The file is parsed locally. Check that the extracted text is what you expect before starting AI translation. Sign-in is required for AI translation.
Step 2: Review request count and start
The page shows how many requests will be sent. Text is split into parts of at most 500 characters. When you start, each successful request uses one daily free translation or 1 credit. You can stop and resume within the session.
Step 3: Copy, download, or reset
After translation, review the XHTML output. Use copy or download to keep the result, example to load sample snippets, and reset to clear the form. Check tags, links, and terms manually; exact visual layout is not guaranteed.
Try this tool
Upload an unencrypted EPUB up to 5MB, review the request count, then translate XHTML text while keeping tags and chapter structure. Sign-in is required for AI translation.