Remove metadata from a Word document

A Word file carries more than its text: who wrote it, who saved it last, your company, how long it was edited, the full path of the template on your computer, custom fields such as a client or matter number, and the GPS position inside photos you pasted in. This tool removes those in your browser, then re-opens the cleaned file and scans it again so you can see they are gone.

• Processed in your browser, no upload. The document never leaves your device.
• Re-scan of the cleaned file. The checks under the download come from opening the new file again.
• Text and layout unchanged. Comments, tracked changes and hidden text are reported, not deleted.

What Word stores in a .docx

A .docx is a ZIP of XML files. The text is in word/document.xml; most of the metadata sits beside it, where you never see it while editing. This is what the tool looks for and what it does with each item.

How to remove metadata from a Word document

1. Add the .docx. Drop the Word file on the tool or pick it. It is read in this browser tab and is not uploaded.
2. Inspect. The Before list shows the author, last modified by, company, dates, template path, custom properties, thumbnail and photo data the file carries.
3. Press Clean this file. The tool clears the document properties, removes custom properties, the thumbnail and the template path, and strips EXIF and GPS from pasted JPEG, PNG and WebP pictures without re-encoding them.
4. Read the re-scan. The cleaned file is re-opened, every XML part is parsed, and it is scanned again. The page lists those checks and any comments, tracked changes or hidden text still in the document.
5. Download the -cleaned.docx. Send that copy. Your original file is not changed.

The download keeps the same format: a .docx comes back as a .docx named yourfile-cleaned.docx. If a check fails, you get a message and no file. A file the tool cannot handle safely is never returned half cleaned.

What the re-scan checks

Before the download button appears, the tool runs four checks on the exact bytes you will download:

• The cleaned file is re-opened as a ZIP and every part's checksum is verified.
• Every XML part is parsed by the tool, then again by your browser's own XML parser, and must read at least as well as in the original.
• No relationship or content type is left pointing to a missing part, so Word has nothing new to repair.
• The file is scanned again with the same inspector. If any removed field is still there, the file is not offered.

We also ran the cleaner on 530 Office files from the Apache POI test data (118 Word, 330 Excel and 82 PowerPoint), saved by Word, Excel and PowerPoint for Windows, Mac and the web, LibreOffice, Apache POI and other programs. Every cleaned file re-opened with valid ZIP checksums, and no XML part that was well-formed in the original became malformed. LibreOffice opened 526 of the originals; all 526 cleaned copies of those opened too and printed to the same number of pages. The same files run through this page in Chrome gave byte-for-byte the same output. Another 55 files in that test data were refused with a message instead of being cleaned: damaged, password-protected, macro-enabled or not really the format their name says.

Word's Document Inspector, and where it stops

On Windows, Word has a built-in cleaner: File > Info > Check for Issues > Inspect Document. Microsoft's page Remove hidden data and personal information by inspecting documents lists what it finds: comments, revisions and versions; document properties and personal information, including the template name; headers, footers and watermarks; hidden text; custom XML data; and invisible content. It is a good tool, and it does things this page does not, such as deleting comments and accepting tracked changes.

Word for Mac has no Document Inspector. Its option is under Review > Protect > Protect Document: tick Remove personal information from this file on save, then save. An answer on Microsoft Q&A describes what it does: the names on comments and tracked changes become “Author”, and the summary properties are cleared except Title and Subject.

A good order for a sensitive file: delete comments and accept tracked changes in Word, save, then clean the saved copy here and check the re-scan. Use the Document Inspector too if you want headers, footers or invisible objects checked; this page does not look at those.

Remove the author from a Word document

If the author's name is the only thing you need gone: in Word for Windows open File > Info, find Related People > Author, right-click the name and choose Remove Person, then save. Word still keeps Last modified by, the company, the template path and the editing time, and it writes your name as Last modified by when you save. The tool clears all of them at once, after your last save.

Comments, tracked changes and hidden text

These are part of the document's content, so the tool reports them instead of deleting them. If it finds any, the result shows a notice such as “This document still has 2 tracked changes by Bob Editor. Accept or reject them in Word before sharing.”

• Comments: Review > Delete > Delete All Comments in Document.
• Tracked changes: Review > Accept > Accept All Changes and Stop Tracking (or reject the ones you do not want).
• Hidden text: Home > Show/Hide ¶ shows it with a dotted underline; select and delete what should not be shared.

Excel and PowerPoint files

The same tool accepts .xlsx and .pptx. It removes their document properties, custom properties and thumbnail, and in Excel the folder the workbook was last saved in (stored in the workbook as an absolute path). Cell comments, threaded comments, hidden sheets and speaker notes are reported, not removed.

Files the tool refuses

• .doc (Word 97-2003): a different, binary format. Open it in Word and save as .docx.
• Password-protected files: the contents are encrypted and cannot be read without the password.
• .docm and other macro-enabled files: save a copy as .docx, which drops the macros.
• Digitally signed documents: any change would break the signature.
• OpenDocument (.odt): save a copy as .docx first.

In each case you get a message saying why, and nothing is downloaded.

Before you send it

• Clean after your last edit. Saving in Word writes Last modified by and the dates again.
• Check the file name. Draft_v3_JSmith.docx says as much as the author field.
• Read the notices. A clean property list does not mean the comments are gone.
• Headers, footers and the text itself are yours to review; the tool does not change content.

Images and PDFs too

Photos you paste into a document have their own metadata before they get there. The image metadata remover handles JPEG, PNG, WebP, HEIC and more, the PDF metadata remover cleans the PDF you export from Word, and the guide by file type covers each format.

What a camera writes into a photo, GPS included, is explained on remove EXIF data.