Convert Word to HTML

Turn a .docx Word document into a clean, styled, standalone HTML file in your browser. No upload, no account, no server round trip.

🌐 Español

Drop your files here (.docx)

🔒 Private by design: your files are processed locally in your browser and never uploaded to any server.

Written and maintained by Last updated

What the converter reads out of a .docx, and what it quietly drops

A .docx is a zip archive full of XML parts, and this tool opens it with mammoth.js and asks only for the semantic layer. Word’s Heading 1 through Heading 6 styles become real <h1> to <h6> elements. Bold becomes <strong>, italic becomes <em>, strikethrough becomes <s>, superscript becomes <sup>. A hyperlink becomes a genuine <a href> with its target resolved through the document’s relationship part. Bulleted and numbered paragraphs are rebuilt into nested <ul> and <ol> lists, and the default style map covers five levels of nesting before deeper items fall back to plain paragraphs. Every Word table becomes <table>, <tr> and <td> markup, and a row marked “Repeat as header row” in Word comes across as a real <thead> of <th> cells.

The omissions matter just as much. Underline has no entry in that default style map, so it vanishes while the text stays. Highlight colour, font colour and font size go the same way. Headers and footers live in separate XML parts that this pipeline never opens, which is also why page numbers cannot come across: there are no pages in an HTML file to number. Text boxes, columns and exact spacing are layout, and layout is precisely what you are trading away in exchange for markup you can restyle.

Footnotes survive, but they move to the bottom

Run a footnoted document through and you will find each marker rendered as a small superscript link, plus a numbered list appended after the last paragraph holding every footnote’s text with an arrow link back to its citation. Nothing is lost, but the reading experience changes: what was a note at the foot of page three is now item three of a list at the end. For a contract or an academic draft that is usually fine. For a document where the notes were doing heavy lifting, it is worth knowing before you paste the result into a wiki.

One stylesheet is baked into every file this tool writes

The download is not a bare fragment. The converted body is wrapped in a full document with a doctype, a UTF-8 charset and an inline stylesheet, so the file opens correctly straight from your downloads folder. That stylesheet sets a 40px body margin, a 16px system sans-serif face with a 1.6 line height, near-black text, images capped at the container width, and tables with collapsed one-pixel borders and real cell padding. It is deliberately plain. If you are pasting into a CMS you will strip it anyway, and if you are not, plain beats the mso- attributes, empty spans and smart-quote entities that copying and pasting out of Word would have given you.

Ten files in, one bad file, nothing out

The drop zone takes several .docx files at once and produces one .html download per document, named from the original. What it does not do is recover from a broken member of the batch. There is no per-file error handling in the conversion loop, so a corrupt or mislabelled document aborts the whole run and no downloads appear at all. If a file is suspect, convert it by itself. A .doc renamed to .docx is the classic offender: the extension check lets it through the drop zone, and the zip reader then fails on it.

Four steps from drop zone to .html download

  1. Drop your .docx files into the box above, or click Choose files to pick them.
  2. Check the list that appears, then click Convert Word to HTML.
  3. Click the Download link for each finished file, one per document.
  4. Click Process another if you want to start a fresh batch.

Where the HTML route beats the PDF route

If your reason for converting is that you want the words, this is the tool to use rather than Convert Word to PDF. That converter shares the same first half of this pipeline and then rasterizes the rendered page into images, so its text stops being text. Here the markup is the deliverable and every word stays selectable, searchable and editable. If you want plain-text structure instead of tags, Convert Word to Markdown runs this same conversion and then rewrites the HTML as Markdown. Going the other way, Markdown to HTML Converter takes pasted Markdown, and Convert HTML to PDF will take the .html file you just downloaded and paginate it. The rest of the writing and formatting tools live on the text tools hub.

Frequently asked questions

Does underlined text stay underlined in the HTML?

No, and this one surprises people. The library behind this converter has no underline entry in its default style map, so an underlined run arrives in the HTML as ordinary text with the underline gone. Bold, italic, strikethrough and superscript are all mapped and do survive. If the underline is carrying meaning in your document, re-apply it in the HTML afterwards or use bold in Word instead.

What happens to footnotes when I convert?

They are kept, but they move. Each footnote marker in the body becomes a small superscript link, and all the footnote texts are gathered into a numbered list appended to the very end of the HTML, each with an arrow link back to where it was cited. That is the only shape HTML has for footnotes without a layout engine, so a document with dozens of them reads more like a reference list than a printed page.

What exactly is in the file I download?

A complete standalone HTML document with a doctype, a UTF-8 charset declaration, an inline stylesheet in the head and the converted body markup. There are no external CSS files, no fonts to fetch and no scripts, so you can open it offline, email it, or paste the body into a CMS. Any images from the Word file are inlined as base64 data URIs, which is why a picture-heavy document produces a large .html file.

Can I convert twenty documents in one go?

Yes, the drop zone accepts several .docx files at once and you get a separate download link per document, each named after its original file. One caveat worth planning around, the run has no per-file recovery, so if a single document in the batch turns out to be unreadable the whole run stops and none of the files become downloadable. Convert a suspect file on its own first.

I dropped a .doc file and it was refused before I clicked anything.

That is the drop zone checking the extension, not the converter failing. Only .docx is accepted, and the shell tells you which file it rejected and which extension it wants. The legacy binary .doc format from Word 2003 and earlier is a completely different structure that this pipeline cannot open. Open it in Word, LibreOffice or Google Docs and save a .docx copy first.

Could a booby-trapped Word file put a script into the HTML?

Not through its text. Every piece of text and every attribute value the converter writes goes through an escaping function first, so a paragraph that literally reads with angle brackets around the word script comes out as escaped entities and renders as visible text, never as live markup. The page you are on also never renders the result, it only hands you a file to download. One caveat is that hyperlink targets are copied from the document as they stand, with no scheme filtering, so check the links before you open or republish an .html file converted from a document you did not write.

Related tools