@markitdownjs/docx

DOCX (Word) to AST converter for MarkItDownJS. Parses headings, paragraphs, lists, tables, inline formatting, and document metadata from .docx files.

Install

npm install @markitdownjs/docx

Usage

import { MarkItDown } from "@markitdownjs/core";
import { DocxConverter } from "@markitdownjs/docx";

const parser = new MarkItDown();
parser.registerConverter(new DocxConverter());

const result = await parser.convert({
  source: docxBuffer,
  mimeType: "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
});
console.log(result.markdown);

Key Exports

| Export | Description | |---|---| | DocxConverter | Converter plugin — register with MarkItDown |

What Gets Extracted

• Headings (mapped from Word heading styles h1–h6)
• Paragraphs with bold, italic, underline, and strikethrough inline formatting
• Ordered and unordered lists (including nested)
• Tables with header row detection
• Document core properties: title, author, description, created/modified dates

Accepted MIME Types

• application/vnd.openxmlformats-officedocument.wordprocessingml.document
• application/msword (.doc files are not supported — convert to .docx first)

Part of the MarkItDownJS monorepo.