Text & HTML to EPUB Converter — Browser-Only, Private

Turn any text or HTML into a Kindle-ready EPUB without sending files to a server. Paste content directly, or pick a .txt / .html file, set title and author, optionally add a cover image, and download the EPUB.

Plain .txt files only in TXT mode; .html in HTML mode. Loaded into the textarea below.
In TXT mode: blank lines separate paragraphs; lines starting with # start new chapters.
Book metadata
Cover image (optional)
JPEG, PNG, or WebP. Around 1600 px on the longest side works well for Kindle.
Waiting for content…

How to make an EPUB from a text file

  1. Paste the text or load a file. When you upload a file, the tool looks at it and flips the source format for you: anything that contains tags like <p>, <div> or a heading is treated as HTML, everything else as plain text.
  2. Mark the chapters. In plain text, put # Chapter One on its own line wherever a chapter starts. Text before the first marker becomes an opening section that appears as "Chapter 1" in the contents.
  3. Fill in the metadata. If you leave the title empty, the uploaded file's name is used, and pasted text with no title is called "Converted".
  4. Add a cover if you have one, then press Convert and Download. The result is named like My_Book.txt-3.epub, showing the source mode and the EPUB version.

Getting HTML chapters right

Chapters split on <h1> elements that sit directly inside <body>. Pages saved from a website or a blog editor often wrap everything in an <article> or <div>; in that case the whole document becomes one long chapter. Strip the wrapper, or move the H1s to the top level, before converting. Every <style> block in the source is copied into each chapter, so a single stylesheet at the top styles the whole book. Images referenced by <img src> are not downloaded into the EPUB, so this tool works best for text-first books; for a picture book, build it with the images to EPUB converter instead.

Good uses and Kindle tips

Typical jobs are turning public-domain texts, fan fiction, long newsletters, meeting notes or a draft manuscript into something you can read comfortably on an e-reader with adjustable font size. For Kindle, send the EPUB with Send to Kindle; a text-only book is tiny, so email delivery is fine. For Kobo, copy it to the device over USB. If you later want a different cover or need to fix the author name, open the book in Change EPUB Cover. If your source is a PDF rather than text, the PDF to Kindle converter extracts the text for you.

Right-to-left page order is enabled automatically for Arabic, Hebrew and Farsi when you build EPUB 3. Chinese, Japanese and Korean books are labelled with their language but keep left-to-right page turns.

Frequently Asked Questions

No. Everything runs in your browser — content never leaves your device.

  • TXT — plain text. Paragraphs are split by blank lines; no formatting.
  • HTML — paste or upload HTML. Headings, lists, links, and inline emphasis are preserved.

Yes. Download the EPUB and use Amazon's Send to Kindle (web, desktop, or email). Amazon converts it to a Kindle format on delivery.

EPUB 3 is recommended for modern readers and Send to Kindle. EPUB 2 is for older devices or strict workflows.

In HTML mode, use <h1> headings — each H1 starts a new chapter automatically. In TXT mode, lines like # Chapter Title on their own start a new chapter. The heading text becomes the chapter's name in the e-reader's table of contents.

JPEG, PNG, and WebP. Recommended size: around 1600 px on the longest side for sharp Kindle display without bloating file size.

Pick Arabic, Hebrew, or Farsi to enable RTL page progression in EPUB 3.

Inline styles and <style> blocks inside the HTML are kept. External CSS files (<link rel="stylesheet">) are not bundled — paste their content as a <style> block instead. Scripts and inline event handlers are removed, since e-readers don't run them.

EPUB renderers (e-readers, Kindle apps) only support a subset of CSS, and reflow text to fit the screen. Use relative units (em, %) and avoid fixed-pixel layouts for best results.

No. In TXT mode, lines that follow each other are joined into one paragraph and only a blank line ends it, so hard-wrapped files such as Project Gutenberg downloads reflow cleanly on any screen size. The flip side is that poems or addresses lose their single line breaks; switch to HTML mode and use <br> where a line must end.

Uploaded files are read as UTF-8. A .txt saved in an older encoding such as Windows-1252 or Latin-1 turns characters like é, ñ or curly quotes into garbage. Re-save the file as UTF-8 in your text editor (Notepad, TextEdit and VS Code all offer it in Save As) and load it again.

If an editor, client or platform might run one, yes: real writers get flagged, and it is better to see it first. Our partner ScanForAI's writers page gives a colour map of the doubted sentences, the reasons, and a top-fixes list, and you can save a share link as a record. Scan chapter by chapter: the limit is 20,000 characters per scan without an account and 50,000 with a free one. Background: why detectors flag human writing.

Only the chapter marker is understood: a line that starts with # and a space. Other Markdown such as **bold**, ## subheadings or lists stays as literal characters. Convert Markdown to HTML first and use HTML mode if you need that formatting.