Remove Metadata From a PDF Before You Share It
Removing PDF metadata means deleting the file's self-describing records: the Info dictionary entries (Title, Author, Subject, Keywords, creator and producer stamps) and the XMP metadata stream that authoring applications embed. This tool blanks the Info dictionary and deletes the XMP stream in your browser, then hands back a cleaned copy to download. It does not remove hidden page content such as cropped or covered text, and it never uploads your file.
How it works
A PDF stores descriptive metadata in two separate places. The older one is the document information dictionary, referenced from the trailer's /Info key: a flat set of strings such as /Author, /Creator, /Producer, /CreationDate, and /ModDate. The newer one is an XMP packet — an XML document following Adobe's Extensible Metadata Platform schema — embedded as a stream object and referenced from the document catalog's /Metadata key. Office suites, scanners, and PDF printers routinely fill both, and they commonly record things you may not want to circulate: the operating-system username of the author, the source document's draft filename, and exact creation timestamps.
When you drop a file on this tool, pdf-lib parses the object tree, overwrites every standard Info entry with an empty string, and removes the catalog's reference to the XMP metadata stream so the packet is dropped from the output. The document is then fully re-serialized. A full rewrite matters here: PDFs saved incrementally keep their previous generations of objects inside the same file, so merely marking fields empty in an incremental save would leave the old strings recoverable with a text editor. The rewrite emits only the current objects, leaving no earlier revision to mine.
Understand the boundary of this operation: it cleans metadata about the document, not content inside the pages. Text that has been covered with a black rectangle, moved outside the crop box, or set in white-on-white is page content and survives untouched — it can still be selected, copied, and searched. If that is your threat model, use Redact PDF, which destroys the underlying text by flattening the affected pages. Embedded file attachments and annotation contents are likewise separate features with their own data.
The processing loop never leaves your machine: the file is read with the browser's FileReader API, transformed in memory, and written back out through a download link. No network request carries any part of the document.
Worked example: cleaning a CV before applying
Take a two-page CV, 412 KB, exported from LibreOffice Writer. Inspecting its properties shows Author: familypc (the shared computer's account name), Creator: Writer, Producer: LibreOffice 24.2, CreationDate: 2024-11-03 23:47 — a middle-of-the-night timestamp — plus a 3.4 KB XMP packet repeating it all, padded with the 2 KB of whitespace the XMP specification recommends so editors can update packets in place.
After a pass through this tool, the output is 407 KB: the XMP stream's 3.4 KB is gone and the rewrite trimmed some slack. Every properties panel now shows blank fields, and the late-night timestamp that hinted the CV was rushed is no longer part of the story the file tells. Elapsed time in the browser: under two seconds.
Frequently asked questions
What metadata does a PDF actually contain?
Most PDFs carry an Info dictionary with Title, Author, Subject, Keywords, Creator, Producer, and creation and modification dates, plus an XMP metadata stream holding the same facts as XML. Files exported from office software often record login usernames and original draft filenames in these fields.
Does removing metadata also remove hidden or redacted text?
No. Metadata removal only clears the descriptive records about the file, not content inside the pages. Text hidden under black boxes or outside the crop box is still in the file and still copyable, so use the Redact PDF tool to destroy it.
Will the PDF look different after stripping metadata?
No. Pages, fonts, images, and layout are untouched because metadata lives outside the page tree. The only visible difference is that reader panels show empty document properties, and the browser tab falls back to the filename.
Why is my PDF slightly smaller after removing metadata?
XMP packets are XML and often include kilobytes of padding whitespace by design, so deleting the stream saves real bytes. The full rewrite also drops old object revisions left behind by incremental saves, which can shrink the file further.
Can someone recover metadata after it has been removed this way?
Not from the downloaded file: the tool performs a complete rewrite, so old values are not retained in earlier revisions inside the file. The original file on your disk still has its metadata, so share only the cleaned copy.
Is it safe to upload a confidential PDF to this page?
Nothing is uploaded at all. The file is opened with JavaScript inside your browser tab, processed in memory on your device, and saved back to your Downloads folder. You can disconnect from the internet after the page loads and the tool keeps working.