Remove metadata from Word documents
A .docx is a zip full of XML, and some of that XML is about you: the author, the last person who edited it, the company and template it came from, comment authors, save-revision IDs and the exact edit times. This tool reads those fields, then removes them in your browser — the document text, styles and images stay as they are.
- 100% in-browser
- No re-save
- .docx
- No account
Clean a Word document
Up to 5 documents · 60 MB each · .docx
Scanned and cleaned locally · no upload · no re-save
What a Word file remembers
Open a .docx as a zip and you find a small set of metadata parts next to the document body. None of them are visible when you read the page, and most of them survive "Save as" and sharing.
Document properties
Author, last modified by, title, subject, keywords, category, revision number, created and modified times, last printed.
Application properties
Company, manager, template name, total editing time, the app and version that wrote the file.
Comments and people
Every comment, its author and initials, and the people list Word keeps for collaboration.
Revision IDs
rsid values stamped on paragraphs and runs across saves — a fingerprint of editing sessions that can link documents back to each other.
What actually changes
The cleaner rewrites the metadata parts of the OOXML package: it blanks document and application properties, drops custom properties and the preview thumbnail, removes comments and their markers, and strips rsid and revision author/date attributes. The body, styles, headers and media bytes are not re-saved.
| Part of the file | Original | After cleaning |
|---|---|---|
| Author / last modified by | Alice Chen · Bob Li | absent |
| Company / manager | Acme Corp · Carol Wu | absent |
| Template / edit time | Normal.dotm · 142 min | absent |
| Created / modified / printed | 2026-10-01 · 2026-10-28 | absent |
| Comments, authors, people | 2 comments · Bob Li | removed |
| Revision IDs (rsid) | 3 identifiers | removed |
How it works
Read the package
The file is unzipped in memory and the document, property, comment and relationship parts are read as XML. Encrypted files and .docm are rejected up front.
Remove only metadata
Properties are blanked, custom properties and the preview thumbnail are dropped, comments and their markers are removed, and rsid / revision author / date attributes are stripped across the Word XML parts.
Re-zip and verify
The package is rebuilt and scanned again. If any personal field survives, the download is refused instead of handing you a file that only looks clean.
What this does not do
- Tracked changes stay. Insertions and deletions are part of the document content; the tool strips their author and date, but does not accept or reject them. Do that in Word if the change history itself must go.
- Text, headers and filenames are untouched. If your name appears in the document body, a header or the file name, it stays there.
- Embedded images and attachments keep their own metadata. Photos inside the document are not re-encoded; use the image cleaner on those files separately.
- Custom XML data parts are not audited. Content controls bound to custom data keep their data store.
- .docm, .dotx and encrypted files are rejected. The tool only handles standard .docx packages it can read without a password.
- Not a guarantee of anonymity. Removing file metadata is one layer; writing style, tracked changes in earlier copies and platform logs are others.
Questions people ask
Are my Word documents uploaded to a server?
No. The file is read, cleaned and downloaded in this tab. Open DevTools, switch to the Network panel, clean a document and confirm that no request carries the file. Closing the tab discards it.
Will Word still open the cleaned file?
It should. The tool rewrites the same OOXML package: it removes or blanks metadata parts and attributes, keeps the document body and media, and re-scans the output before offering it. Test the copy in Word before sending anything important.
What happens to comments and tracked changes?
Comments are removed completely, including the comment parts, the people list and the inline comment markers. Tracked changes are kept as content because deleting them would change what the document says, but their author and date attributes are stripped. Accept or reject changes in Word first if you want them gone.
Does it remove my name from the text of the document?
No. It removes file-level metadata, not words on the page. If your name appears in the body, a header or a filename, it stays there. Search the document in Word if that matters.
Does it work on .docm or password-protected files?
No. Macro-enabled .docm files are rejected so that macros are not silently altered, and encrypted or password-protected documents cannot be read in the browser. Save a .docx copy and remove the password in Word first.
Is it different from Word's own Inspect Document?
They overlap, but this tool runs in any browser without installing Office, never uploads the file, and re-scans the cleaned output before download. Word's inspector is the better choice when you can use Word and want a full document check.
Other Docfy tools
Send the document, not the author.
Strip Word's hidden properties, comments and revision IDs in the browser — text and layout untouched.