Word documents have structure you don't want to throw away when converting to Markdown — the heading hierarchy you carefully built, the nested bullet lists, the bold/italic emphasis, the tables. mdstill preserves these when converting .docx and .doc to GitHub-flavored Markdown, so you keep the structure you already have.
Max file size: 20MB ·
Word's heading styles (Heading 1, Heading 2, etc.) map directly to Markdown H1–H6. Bold and italic character formatting become **bold** and *italic*. Ordered and unordered lists — including deeply nested ones — are preserved with correct indentation. Tables convert to GFM Markdown with pipes and header separators. Hyperlinks keep their URLs in [text](url) form.
Markdown is intentionally simpler than Word. Specific fonts, font sizes, and colors are dropped — they don't exist in the Markdown spec. Text-box overlays, columns, and page-specific layout (headers, footers, page breaks) are flattened into the main document flow. If you relied on these for meaning, you'll want to restructure before relying on the output.
Old-school binary .doc files (pre-2007 Word) go through a LibreOffice conversion step first to normalize to .docx, then through the regular pipeline. This all happens server-side in one API call — you just upload the .doc and get Markdown back. The only practical difference is that .doc files take a second or two longer than native .docx.
Moving existing documentation from Word into a static site generator (Hugo, Docusaurus, Astro). Pulling Word-based knowledge bases into wikis or Obsidian vaults. Preparing Word source material for LLM context windows — Markdown uses 40–60% fewer tokens than raw .docx extraction. Migrating legacy project docs to GitHub-hosted Markdown.