Convert a Word document to PDF

Converting a document to PDF is what you do at the end: the file is finished and needs to be readable everywhere, unchanged, by someone who may not have Word.

That final step is also the moment the document is most sensitive. A CV, a signed letter, an invoice, a contract, a medical form. It is a strange habit of the web that this is exactly when people are asked to upload the file to a stranger’s server. This conversion happens in your browser. The .docx is unpacked, read, and typeset into a PDF without anything being transmitted.

A .docx is a zip of XML, which is why no upload is necessary and no heavy conversion engine is either. The interesting part is not unzipping it but interpreting it. Word does not store a “bulleted list”: it stores a paragraph carrying a numbering reference, and whether that reference means a bullet or a number lives in a separate part of the archive. Following that chain is the difference between a PDF that keeps its lists and one where every item quietly becomes an ordinary line. Headings work the same way, through named styles rather than through font size, which is why a document that only has larger bold text will convert as plain paragraphs.

What carries over is text, heading levels, lists, tables drawn as a grid, PNG and JPEG pictures, the text of text boxes, and footnotes, which are printed after the body. Hidden text stays hidden. What does not carry over is headers, footers, and anything positioned rather than flowed, and the result says when the document had them. That is a real limit and worth stating before you convert rather than after. If your document is a letter, a CV, an essay, notes, or a report, the result will be what you expect. If it is a designed brochure or built around a table, printing to PDF from Word itself preserves the layout properly and this will not.

One more limit. Latin text, curly quotes, and dashes use the standard PDF fonts, and accented, Greek, Cyrillic, Arabic, and Hindi text is drawn with fonts this site serves, loaded only when a document needs them. Chinese, Japanese, and Korean are not covered yet. Rather than fail or silently drop those characters, they are replaced and the number replaced is reported with the result, so you know before sending the file rather than after.

The older .doc format is not supported. Despite the similar name it is a completely different, binary format from an earlier era of Word. Open it in Word or LibreOffice and save it as .docx, then convert.