DOCX to Markdown Converter

How to use this tool

1. Choose a .docx file with "Choose file", or drop it onto the box. You can also press "Try an example".
2. Read the Markdown in the Markdown box. It appears as soon as the file is read.
3. Tick "Embed images" if you want pictures kept inside the Markdown; the result updates straight away.
4. Press "Copy Markdown", or download the result as a .md file named after your document.

What happens to each part of a Word document

Word stores a document as styled paragraphs. The converter keeps the structure those styles describe, such as headings, lists, links and tables, and lets go of the look, such as fonts, colours and spacing, because Markdown has nowhere to put it. Each result below was produced by the converter itself.

# Results

## Detail

**Done** and *nearly*

~~draft~~ final

[CommonMark](https://commonmark.org/)

- milk
- eggs

1. Plan
   1. Draft
2. Write

| Item | Qty |
| ---- | --- |
| Pens | 4   |

> Measure twice

Claim[^1]

[^1]: Source.

one\
two

Why convert Word documents to Markdown

Markdown is plain text with a few symbols for structure, which makes it the common language of documentation sites, README files, wikis, note apps such as Obsidian, and static site generators. It also suits AI tools: a language model reads a heading marked with # and a table drawn with pipes far more reliably than it reads the XML inside a .docx, so Markdown is the usual format for feeding documents into a chat, a custom GPT or a retrieval pipeline.

Copying from Word and pasting into a Markdown editor loses most of that structure: headings arrive as plain lines, numbered lists lose their numbers and tables collapse into tab-separated text. Reading the .docx itself avoids that, because the converter sees which paragraphs carry a Heading style, which belong to a list and which cells make up a table.

How it works

A .docx file is a ZIP archive of XML files. The page opens it with mammoth, an open-source library that maps Word's built-in styles to their meaning: Heading 1 becomes a top-level heading, a list paragraph becomes a list item, and a run marked bold becomes strong text. That structure is then written out as GitHub Flavored Markdown, with special characters escaped so that an asterisk you typed in Word stays an asterisk instead of turning into italics.

Because the mapping follows styles, documents that use Word's own Heading styles convert best. A line that only looks like a heading because it was made big and bold by hand is still a normal paragraph to Word, so it comes through as bold text. Applying the Heading styles in Word before converting fixes that.

Private by design

The file never leaves your device. It is read into this page with your browser's file reader, converted by code that is already loaded, and the Markdown is shown in the box above. There is no upload, no account and no copy on a server, and the tool keeps working if you go offline once the page has loaded.

Limits

• Only .docx files. Old .doc files, PDFs and .odt files are not read.
• Fonts, colours, highlighting, underlining, page layout, headers, footers and text boxes are not converted. Markdown has no equivalent for most of them.
• Tables become pipe tables with the first row as the header. Markdown cannot merge cells, so a merged cell's text goes in its first cell and the cells it covered are left empty. Lists or several paragraphs inside a cell are joined onto one line, because a Markdown table cell holds a single line.
• Numbered lists always start at 1, even if the Word list started at another number.
• Comments are left out, and tracked changes are not shown as changes; accept or reject them in Word first.
• Equations, charts and SmartArt are not converted. Images are left out unless you embed them.