Word to Markdown, without losing the structure
A .docx already knows what its own headings and lists are, which makes it the easiest thing here to convert well. The interesting question is not whether it will work — it will — but which parts of a Word document are structure worth keeping and which are only appearance.
How to do it
Check that your headings are real headings
Markdown can only keep a heading that Word knows is a heading. A line styled with Heading 1 becomes #; a line that is merely 18pt and bold becomes an ordinary paragraph, because as far as the file is concerned that is all it is. If the outline pane in Word shows your document's structure, the conversion will keep it.
Drop the file in and convert
There is nothing to configure. A .docx carries its own text, so no document language and no OCR are involved, and fast mode is the right choice.
Read the result, not the layout
The output will not look like your document, and it is not meant to. Fonts, colours, margins, page breaks, columns and spacing are presentation; Markdown keeps meaning. What survives is the outline, the lists, the tables, the links and the emphasis.
Tidy up what extraction left behind
Open the review panel on the file's row for the optional cleanup: joining words split across a line break, and reducing runs of blank lines. Each one previews first, tells you how many passages it would change, and is a single undo step once applied.
A worked example
A short brief with two heading levels, a bulleted list and one external link — the four things worth carrying across.
What goes in
Northwind Ferry Terminal [Heading 1] Project brief, second revision [Heading 2] Scope [Heading 2] • Replace the covered walkway. • Add a second ticket hall entrance. Full guidance is at example.org/access-guidance [hyperlink]
What comes out
# Northwind Ferry Terminal ## Project brief, second revision ## Scope - Replace the covered walkway. - Add a second ticket hall entrance. Full guidance is at [example.org/access-guidance](https://example.org/access-guidance).
The converter's “Text document” sample is exactly this kind of file, if you want to see the output before using your own.
What it will not do
- Appearance does not survive, by design: fonts, colours, margins, page breaks, columns, text boxes and exact spacing are all presentation rather than meaning.
- Tracked changes and comments are not carried across. Accept or reject changes in Word first, or you may not get the text you expect.
- Headers, footers and page numbering belong to the printed page and have no equivalent in Markdown.
- An embedded object — a chart, an equation, a linked spreadsheet — is not converted. Its text does not appear.
- A password-protected .docx needs its password, which is used in this tab and never stored or sent.
- A document with macros is refused before any of it is read. That is a security decision and it is not configurable.
- Older .doc files, from before the .docx format, are not supported. Save as .docx first.
Questions
- Why does my result look nothing like my document?
- Because Markdown has no fonts, no margins and no pages. It keeps what a heading is, not what a heading looks like. That is the point when the destination is an AI system, a wiki or a static site.
- Do my tables survive?
- Ordinary tables do, as Markdown tables. Merged cells do not have a Markdown equivalent and are squared off, so check any table where cells span rows or columns.
- What about images in the document?
- The single-file download includes them in a package alongside the Markdown. The split export is text only, and its index says so rather than leaving you links to files that are not there.
- Is my document uploaded to convert it?
- No. It is read by code running in this tab, on your device. Nothing about it is sent anywhere, and there is no account.
- Can I convert a folder of documents at once?
- You can select several files and convert them as a batch. Each row gets its own copy button and its own review, and one file failing does not take the rest down with it.
Where your file goes
Nowhere. Reading, converting, cleaning and counting all happen in this browser tab, on your own device. The file, its name and its text are never uploaded to a server, and there is no account to create.
One thing does use the network, and it is worth being precise about: reading a scanned page needs an OCR model, which downloads from this site the first time you convert an image or a scan. That is a program arriving on your device, not your document leaving it. After it has been fetched once, that conversion works offline.