PDF Metadata Viewer & Editor

about this tool

What a PDF says about itself

Drop a PDF and the panel shows what the file carries: page count and paper size, the version in its %PDF- header, how many times it has been saved (each %%EOF is one revision, so more than one means it was updated in place), and whether it holds an XMP packet. Below that are the eight Info dictionary fields — title, author, subject, keywords, creator, producer, created and modified — each editable, with a list of exactly what your edits will change before you save.

The "what it gives away" panel spells out the identifying parts: the author field is usually the account name of whoever made the file, the creator and producer name the software, and the dates are precise to the second and time zone. The file name is not stored inside a PDF, so renaming a file tells nobody anything.

Dates, exactly as written

PDF dates look like D:20260910143005+05'30', and almost every part after the year is optional. The parser here follows PDF 32000-1 §7.9.4: a bare Z, an hour-only offset, missing apostrophes and truncated forms down to just a year all read correctly, and a date is written back in the same shape it came in. A document made at 14:30 in Delhi keeps saying 14:30 with a +05:30 offset rather than being shifted into your time zone. Editing a date keeps the offset the file used; a date you add takes your computer's. Impossible dates — month 13, 31 February — are refused with the reason.

Removing it

Emptying a field deletes the entry from the Info dictionary rather than storing an empty string, so nothing is left for another tool to read. Ticking the XMP box deletes the packet object itself, not just the reference to it: pdf-lib writes every object it knows about, reachable or not, so unlinking alone would leave the XML sitting in the bytes.

Text is written as UTF-16, so any language survives. Saving rewrites the document in a single pass: pages, text, images and annotations are untouched, but earlier saved revisions inside the file are dropped and the size will differ from the original.

Limits

The XMP packet can be read for a few common fields and removed, but not rewritten — matching one to edited Info fields properly means a full XMP writer. Encrypted PDFs are refused, since their strings cannot be read without the password. Per-page metadata, embedded file attachments, bookmarks and form data are not shown. Everything runs in your browser and nothing is stored between visits.

Metadata hides in image files too — EXIF viewer & remover strips the camera, timestamp and GPS tags a photo carries. To stop a document being edited at all, protect PDF adds a password.

questions

What does a PDF know about me?

The Info dictionary usually holds an author name taken from your account, the program you wrote the document in, the program that produced the PDF, and the creation and modification times down to the second and time zone. The panel lists what your file actually carries.

How do I remove all of it?

Empty every field, then save. The entries are deleted from the file rather than blanked, so nothing is left behind, and the XMP packet goes too when you tick that box.

What is the XMP packet?

A second copy of the metadata in XML, which many programs prefer over the Info dictionary and which often holds more: editing history, document identifiers, the tools that touched the file. This tool can show a little of it and remove it, but it cannot rewrite it, so editing the fields while keeping the packet leaves the file saying two different things.

Are the dates changed to my time zone?

No. Dates are read and written as the exact string in the file, so a document made at 14:30 in Delhi still says 14:30 with a +05:30 offset. Editing a date keeps whatever offset it had; a date you add gets your computer’s.

Does saving change anything else?

The file is rewritten in one pass, so earlier saved revisions inside it are dropped and the byte size will differ. Pages, text, images and annotations are untouched.