IntactFile

"Word found unreadable content"

That prompt means Word opened your file, found the package structure it expected, and hit XML it could not parse. The words are almost always still inside — the problem is the wrapper around them, and fixing the wrapper does not require sending your document anywhere.

Your file never leaves your device — repair runs in your browser. 0 bytes uploaded.

Your data is the big block — usually intact. What breaks is the small index. Repair rebuilds it.

Drop a file here or browseTap to choose a file

ZIP · Office · PDF · video · JPG · PNG · RAR/7z · SQLite — repaired right here in your browser. Nothing is uploaded

0 bytes uploadedFree: 3 repairs/day — up to 500 MB video, 100 MB docs & archives, 50 MB photos. Free download. Single repair €5.90. No account required.

Fix it now

  1. Drop the file
  2. Repair runs locally
  3. Download the result

When Word shows "Word found unreadable content in .docx. Do you want to recover the contents of this document?", it has detected that the file is a valid enough package to attempt, but one of its internal parts is broken. A modern .docx is not a single binary blob — it is a ZIP archive full of XML files. Drop the document above and the tool opens that package in your browser, repairs the container and the malformed parts, and rebuilds a file Word will open cleanly — without the document ever leaving your machine.

A .docx is a ZIP of XML — and that is why it breaks this way

Since Word 2007, .docx uses the Office Open XML (OOXML) format, packaged with the Open Packaging Conventions. Rename one to .zip and you can look inside: you will find [Content_Types].xml, a _rels/.rels map, and the word/ folder holding document.xml (the actual text and paragraph structure), styles.xml, settings.xml, fontTable.xml, a word/media/ folder for images, and word/_rels/document.xml.rels tying it all together.

The heart of the file is word/document.xml. Every paragraph is a <w:p>, every run of text a <w:r> wrapping a <w:t>. XML is strict: one unclosed tag, one stray byte, or a stream that stops mid-element makes the whole part unparseable. Word is unforgiving here by design — rather than silently rendering half a document, it stops and offers recovery. So the error has two common root causes: the ZIP container itself is damaged (a bad copy, a sync that mangled bytes), or one XML part — usually document.xml — is malformed or truncated.

What repair actually gets back

Because the format is layered, repair works at two levels. First the tool treats the file as a ZIP and rebuilds the package structure — fixing the directory, re-reading each stored part, and re-assembling a well-formed archive. Then it repairs the OOXML parts inside: it re-validates document.xml and its siblings, closes structures that were left open, and drops fragments that cannot be salvaged so the remaining text parses cleanly.

In practice this recovers the body text and much of the formatting — paragraph styles, headings, tables and lists that were fully written survive because they are all in document.xml and styles.xml. Embedded images tend to come through well: they live as independent files under word/media/ (PNG, JPEG and so on), so damage to the text XML does not touch them, and a picture whose part is intact is simply carried into the rebuilt package. All of this happens in your browser tab — no Word install, no add-in, no server round-trip.

Where the content genuinely stops

Repair recovers what is present; it cannot reconstruct text that was never written to disk. If document.xml was truncated — the save or copy stopped partway — everything after the break point is gone, and the tool can only return the paragraphs up to that cut. That is the honest ceiling on any recovery of this file, ours or Word's own.

Some richer constructs are also more fragile than plain paragraphs. Complex fields (auto-updating cross-references, some form fields) and tracked changes carry extra markup that, if it lands on the damaged side, may be lost or flattened even when the surrounding text survives. When that happens you generally keep the readable text and lose the mechanism, not the whole document. For anything confidential — contracts, medical notes, unpublished work — the repair staying on-device is the point: you can confirm in the Network tab that zero bytes of the .docx are ever sent.

What this can and can't fix

Can fix

  • A .docx whose ZIP/OPC container was damaged in a bad copy or sync
  • Body text and paragraph structure when document.xml is repairable
  • Styles, headings, tables and lists that were fully written before the damage
  • Embedded images, which live as separate parts under word/media/ and usually survive
  • Files Word will only open through its own 'recover contents' prompt

Can't fix

  • Text past the break when word/document.xml was truncated — those bytes were never saved
  • Tracked changes and complex/auto fields whose markup fell on the damaged side (text may survive, the mechanism may not)
  • Password-encrypted .docx files without the password
  • A file that is 0 bytes or contains no recognizable OOXML parts

If a repair fails, we tell you why (missing data versus broken structure), and you are never charged for a failed repair.

FAQ

Should I click 'Yes' when Word offers to recover the contents?

It is worth trying first — Word's own recovery sometimes succeeds. When it fails, opens a nearly empty document, or drops most of your text, repairing the OOXML package directly usually recovers more, because it rebuilds the ZIP container and the XML parts rather than only reading what Word can already parse.

Will I get my formatting and images back, or just plain text?

Usually much more than plain text. Paragraph styles, headings, tables and lists live in the XML that repair rebuilds, and images are stored as separate files inside the package, so an intact picture is carried straight into the recovered document.

Why does Word call it 'unreadable' when the text is clearly in there?

Word parses the internal XML strictly. A single malformed or truncated part — most often word/document.xml — makes that part unreadable to the parser, so Word refuses the whole file even though the surviving parts are perfectly fine. Repairing the XML lets it open normally.

Is my document uploaded to be repaired?

No. The .docx is read from your disk and rebuilt in your browser; nothing is transmitted. You can watch the Network tab and confirm 0 bytes leave — which matters for contracts, records and drafts you would not want copied to someone else's server.

Related: Repair an Excel workbook · "Excel cannot open the file" · Repair a PDF