Convert Markdown to Word

Markdown to Word is a structure conversion, not a rename

A Markdown file is plain Unicode text. A Word file is a ZIP package containing XML parts, relationships, styles, numbering definitions and any embedded media. Changing .md to .docx changes none of that. This converter reads the Markdown structure and writes a real Office Open XML package: headings become Word heading styles, emphasis becomes run formatting, links become relationships, lists get numbering records, tables become Word tables and code receives a distinct paragraph style.

The package is validated before the download button appears. Its XML must be valid UTF-8, the required parts must exist, the root relationship must point to the main document and internal relationships may not point at missing files. The downloadable JSON receipt lists every part and SHA-256 hash, so verification is evidence rather than a green checkmark with no explanation.

Markdown and DOCX, in brief

Markdown

Full name	Markdown
Extension	.md, .markdown
Format type	Plain text with lightweight formatting
MIME type	text/markdown

DOCX

Full name	Office Open XML Word document
Extension	.docx
Format type	ZIP package of XML document parts and media
MIME type	application/vnd.openxmlformats-officedocument.wordprocessingml.document

What survives, and what needs review

Headings from level one through six, paragraphs, bold and italic text, strike-through, inline code, fenced code, nested ordered and unordered lists, blockquotes, thematic rules, pipe tables and safe web or email links have direct Word representations. PNG and JPEG data URIs can be embedded. Repeated copies of the same image share one media part while every occurrence remains visible in the receipt.

Markdown can also contain things that DOCX does not understand in the same way. YAML front matter is metadata for publishing systems, not document prose. A fenced-code language is a hint for a syntax highlighter, not executable Word metadata. Task checkboxes become static text. Unsupported extensions remain visible as literal content where possible. Each case is named as a review or loss finding, with a count and line when the source makes one available. The full report is never clipped in the downloaded artifact even when its on-screen summary is.

Raw HTML stays inert and remote images stay remote

Markdown permits raw HTML, including tags with event handlers and remote image addresses. Treating that HTML as a live preview would allow a pasted document to run markup or contact a server. Here it becomes literal document text. The converter makes no network request, and a remote image becomes a visible placeholder naming its alt text and source. The receipt records why it was not fetched. This is a deliberate privacy boundary, not an intermittent image error.

If an image must travel inside the Word file, put its bytes in a PNG or JPEG data URI. Relative paths such as images/chart.png cannot resolve from a pasted string because the browser has not been given that file. They remain visible and receive a finding rather than quietly turning into a broken relationship in the DOCX package.

Three honest ways to move Markdown into Google Docs

Google Docs now supports Markdown in three relevant ways. Its official help describes importing and exporting Markdown files and a Paste from Markdown command. The Workspace update announcement explains that the import creates a new Google Doc and preserves Markdown as document structure. That makes Download .md the most inspectable handoff: the exact Markdown remains a file you can keep, diff and import.

Copy raw puts Markdown text on the clipboard for Paste from Markdown. Copy styled instead offers both HTML and plain text, after replacing remote images, for an ordinary formatted paste into Word or Google Docs. Browser clipboard permissions and the destination application decide which flavor is accepted, so these are separate buttons rather than one button claiming universal clipboard magic. See Google’s Markdown help and the Workspace announcement for the destination-side steps.

Page settings belong to Word, not Markdown

Markdown has no page size, margin, header or paper concept. The converter therefore asks only for the presentation facts needed to produce a useful first DOCX: US Letter or A4, three named margin widths and an optional document title. It does not invent typography controls whose effect would be hard to audit. Word styles remain editable after download, which is the right place to change fonts, spacing or a corporate template.

If a fixed final page layout is the real goal, Markdown to PDF shows the page breaks before conversion. If the destination is a website rather than a word processor, Markdown to HTML exposes both rendered output and source. For editing before either conversion, use the Markdown editor; it uses the same styled, raw and download handoff controls shown here.

The file never leaves this tab

File reading, Markdown parsing, image decoding, XML generation, ZIP packaging and validation all run locally in the browser. A large source moves to a background worker so the page remains responsive, but it still stays on the same device. The page has no conversion upload endpoint, no temporary server copy and no deletion schedule. For a technical description of the package, Microsoft publishes an Open XML package overview.