See and Remove the Metadata in a PDF

What a PDF says about you

The details are written automatically, by whichever program made the file, and most people never see them. They are read in a couple of clicks by anyone who receives the document.

• Author is usually taken from the account name on the computer that made it. On a work laptop that's often your full legal name, whether or not you ever typed it.
• Creator and Producer name the software and frequently the exact version — which says what you run, and sometimes what your employer licenses.
• Created and Modified timestamps include a UTC offset, so they leak the time zone you were in, and reveal a document written at 3am or finished after a deadline.
• Document ID is a persistent identifier that survives edits. Two files that look unrelated can be tied to the same original by it.
• Title is very often not what you'd guess: converters commonly set it to the original file's name or its first heading, so final_v3_REAL_use_this_one travels with the document you thought you'd cleaned up.

This matters most in exactly the situations where people are least likely to check — an anonymous complaint, a tip to a journalist, a job application you'd rather not have traced to your current employer's machine, a document filed publicly by a court.

A worked example

You write a reference letter in Word on the office laptop and export it to PDF. The document you send looks clean. Inside, the Author reads your full name as IT configured it years ago, the Producer says Microsoft Word for Microsoft 365, the creation timestamp shows 23:41 with a +01:00 offset, and the Title is Letter template FINAL (Ruth's copy) because that's what the file was called before you renamed it.

Drop it here and all of that is listed on screen. Take the clean copy and every one of those fields is gone from the file — not blanked, not overwritten, gone — while the letter itself is untouched.

The two hiding places, and the mistake most tools make

Metadata lives in two separate places in a PDF, and they don't always agree with each other. The old document information dictionary holds the classic fields. Alongside it sits an XMP packet — a block of XML, added by most modern software, which repeats much of the same information and usually adds more, including the document identifiers. Clearing one and leaving the other is the commonest way a "metadata remover" gives you a false result: the properties box looks empty while the XML underneath still names you.

There's a subtler failure too, and it's worth understanding because it's invisible from the outside. A PDF is a collection of numbered objects with references between them. Removing the reference to the metadata makes it vanish from every properties dialog — but the object is still sitting in the file, and anything that reads the raw bytes still finds it. It is the same mistake as covering a paragraph with a black rectangle and calling it redacted.

This tool deletes both places, and deletes the objects themselves rather than the pointers to them. You can check: open the cleaned file in a text editor and search for your name. It isn't there.

What this does not remove

Being straight about the limits is the only way this is useful. A clean metadata card doesn't make a document anonymous, and these are all content rather than metadata:

• Comments and annotations carry the name of whoever wrote them, right there in the markup. If the tool finds any it lists them, but taking them out would change your document, so it leaves them alone. Delete them in your PDF editor.
• Text under a black box is still text. Covering it draws a rectangle on top; the words remain selectable underneath. The only reliable fix is to remove the whole page.
• Attached files travel inside the PDF with their own names and dates, and are flagged here if present.
• The writing itself. Phrasing, letterhead, and the printer or scanner model baked into a scanned image are all still there.

Frequently asked questions