Remove PDF Metadata

At minimum: title, author, subject, keywords, the software that created it, the software that produced the PDF, and creation and modification timestamps. Beyond that, figures placed from design software carry their own copy of all of this - frequently including the original filename of the artwork - and some tools write private data blocks holding editable versions of the graphics.

We remove the document properties, the XMP metadata stream, metadata dictionaries embedded anywhere in the document, design-software private data blocks, and annotation author names. What we cannot remove is text written inside the compressed image data of a figure itself - some design software stamps its name there. Removing it would require re-encoding the graphic, which would change how your figure looks. When we detect this, we tell you rather than pretend the file is clean. If a document is extremely sensitive, re-export the figures as flat images before placing them.

It should not, and we check. After cleaning, the file is re-parsed and three things are verified: the page count, the number of bookmarks, and that no internal reference points at a deleted object. If any check fails, you receive your original file back untouched. We would rather leave metadata in place than hand you a broken document.

Selected file contents are processed in your browser without being sent to an application upload endpoint. Loading the site, fonts or processing libraries may still require network requests.

Removing metadata and private editing data may reduce file size. The amount depends on the document; a PDF with little metadata may change very little.

Figures that relied on embedded private data will no longer round-trip back into the design software for lossless editing. The PDF displays and prints exactly as before. If you need to keep editing the artwork, work from your original design files rather than from the PDF.

No. An encrypted PDF cannot be modified without the password, and this tool will tell you the file is encrypted rather than failing silently. Remove the protection first, clean the file, then re-apply protection if you need it.

No, and this is an important distinction. If someone drew a black rectangle over text in a PDF editor, the text is usually still there underneath and still selectable. That is a redaction problem, not a metadata problem, and it needs a dedicated redaction tool. This tool cleans metadata only.

Yes. This matters more than it sounds: bookmark titles, form field labels and annotation captions use the same underlying PDF key as the document title. A tool that deletes every title key will silently wipe out your navigation. We identify metadata dictionaries by their signature - a dictionary is only treated as metadata if it carries several metadata fields together - so document structure is left alone, and the bookmark count is verified afterwards.

Drop the cleaned file back into this tool - it should report nothing found. For independent confirmation, run exiftool on the file from a terminal, or open the document properties panel in your PDF reader.

You can process up to 20 files at a time, each up to 50MB. These limits exist only to keep your browser responsive, not to push you towards a paid plan - there is no paid plan.

No. Pages, text, images and layout are untouched. Only metadata and software private data are removed, and the page and bookmark counts are verified to be identical afterwards.