Word files: a DOCX is a ZIP full of XML, and why converting into one is hard
What a DOCX really contains, why PDF to Word loses the layout, and why we can write Word files but not read them. Convert PDF to Word in your browser, no upload.
At a glance
- Type
- Documents
- Extensions
- .docx
- MIME type
- application/vnd.openxmlformats-officedocument.wordprocessingml.document
- Also known as
- DOCX, Microsoft Word, Office Open XML
A .docx file is a ZIP archive. Rename one to .zip, open it, and you will find a folder of XML documents — the text and its structure in one file, styles in another, numbering, relationships, embedded images as separate entries. Microsoft moved to this format, Office Open XML, in 2007, replacing the opaque binary .doc that came before. Being XML in a ZIP is why so many tools can produce a DOCX with no Microsoft code involved, including this site.
Read more
That asymmetry is worth stating plainly: writing a DOCX is tractable, reading one faithfully is not. Producing a document means generating XML for paragraphs, headings and runs — a library can do it. Rendering an existing DOCX exactly as Word would means reproducing decades of layout behaviour, which in practice requires Word or LibreOffice. This is why the conversions here go into DOCX and not out of it.
When the source is a PDF, the difficulty moves to the other side. A PDF stores positioned glyph runs rather than paragraphs, so building a DOCX from one means inferring structure that was never recorded: grouping glyphs into lines by their baselines, guessing paragraph breaks, ranking type sizes to identify headings. Ordinary prose comes through well. Multi-column layouts, tables and floating images do not, because the information needed to reconstruct them is not in the file.
If the PDF is a scan, there is no text at all to work with, and this site reads the characters off the page images with an AI recognition model running on your own machine rather than uploading the file. The result is a Word document you can edit, with the caveat that recognition is not perfect and the layout is reconstructed rather than preserved.
Convert other formats to Word
FAQ
Why does PDF to Word never match the original layout?
Because a PDF stores positioned glyphs rather than structure. Paragraphs, headings and reading order have to be inferred, and a table or two-column layout that was never recorded as such cannot be recovered.
Can I convert Word to PDF here?
Not locally. Rendering a DOCX exactly as Word does requires Word or LibreOffice, which cannot run in a browser tab. Conversions here go into DOCX rather than out of it.
Is a DOCX really a ZIP file?
Yes. Rename it to .zip and you can open it — inside is a folder of XML files describing the text, styles and relationships, plus any embedded images as separate entries.
Does it work on scanned PDFs?
Yes. When a PDF has no text layer, an AI recognition model runs in your browser to read the characters off the page images. Nothing is uploaded.
Are headings and page breaks preserved?
Headings are inferred from type size and page breaks from the original are inserted, so the document keeps its pagination. Both are guesses, and both are occasionally wrong.
Would a server-side converter do better?
On layout, yes — a server running LibreOffice keeps far more of it. It also has to receive your file, and PDFs tend to be contracts, payslips and statements. That trade is the reason this converter works the way it does.