Remove Metadata from PDF

A PDF remembers more than what is printed on its pages. Behind the text sit the author’s name, the program that made the file, the creation and edit dates, XMP properties, an identifier that follows the document around and, sometimes, earlier versions of the pages themselves. A PDF metadata remover deletes those records and leaves the pages exactly as they look. Drop a PDF in the box above to see what it holds and download a clean copy. The file is processed in your browser and never uploaded.

What a PDF remembers about you

Where the author name comes from

People are often surprised to find their name in a file they never typed it into. The fields are usually filled in automatically by whatever made the PDF.

• Word processors copy the name from your user profile into Author and write their own name and version into Creator and Producer.
• Operating systems. Saving a PDF from the Mac print dialog stamps “Quartz PDFContext” and the macOS version as the Producer.
• Browsers. A page printed to PDF from a Chromium browser shows “Skia/PDF” as the Producer and uses the page title as the Title. The Creator can include the browser’s identification string.
• Scanners and copiers commonly record the device model and firmware as Creator and Producer.
• Design tools keep project and software details in XMP and in private application data.

Hidden edit history and failed redactions

Some editors save a change by appending it to the end of the file instead of rewriting the document. The old objects stay inside the file. A viewer shows the newest version, but the previous text, the earlier author and the old dates are still there for anyone who knows where to look. This is why an edited PDF can leak things that are on no page.

The cleaner rewrites the file from scratch and keeps only what the current version of the document points to. In our tests, a PDF with two saved revisions came out holding only the current page text. The earlier text, author and dates were gone.

Why editing the Author field is not enough

A common fix is to open the file in a PDF editor, blank the Author in Document Properties and save. It looks right in the Properties window, and it is not reliable.

• Two places hold the same facts. The Info dictionary and the XMP packet both store author, creator and dates. Clearing one can leave the other.
• Saving may append. Many editors save by adding the change to the end of the file, so the old Info dictionary with the old author stays inside, unused but readable.
• Printing to PDF makes new metadata. Re-printing a PDF through a virtual printer drops some fields and invents others, such as a new Producer and a Creator naming the print driver. Some printers also flatten links and bookmarks.
• Optimising or compressing is not cleaning. A smaller file can still carry every property it started with.

A rewrite that drops everything the current document no longer uses, as this tool does, is the dependable way to be sure.

Who needs to clean PDFs

• Job seekers. A CV exported from an employer’s template can show your current company in the Creator or Author field.
• Academic authors. Double-blind peer review only works if the PDF does not name you, and the Author field is the first place reviewers look.
• Lawyers, HR and finance teams. A document that passed through several hands carries their details and earlier drafts in its saved versions.
• Businesses sending proposals and downloads. Producer and template details expose your tools, your agency and internal project names.
• Public bodies releasing documents. Old versions and author details are a recurring cause of accidental disclosure.
• Anyone sending photographed documents as PDF. The embedded photos can carry GPS and device details.

Scans, OCR and photographed documents

Scanner and copier PDFs usually record the device model and firmware in Creator and Producer, and store each page as a JPEG image. Phone scanning apps do the same and may embed the camera photo with its EXIF, including location. The cleaner removes the document fields and the metadata inside those page images, and keeps the picture and any OCR text layer exactly as they are. The text of a scanned page is visible content, so a name printed on the page stays.

What the PDF metadata remover changes

Removed

• Every field in the Info dictionary, including custom ones such as Company, and the dates.
• XMP metadata on the document, on pages and on images, and application-private data.
• The document ID in the trailer.
• Unused objects: earlier saved versions, old Info dictionaries and orphaned XMP.
• EXIF, GPS and comments inside photos embedded as JPEG. The picture data itself is copied unchanged.

Kept

• Pages, text, fonts, images and vector graphics, exactly as they were.
• Links, bookmarks, form fields and the values typed into them.
• The accessibility structure: tags and the document language. In our tests tagged PDFs kept both.

Not touched

• Anything visible on the pages, including text layers from OCR and names in headers or footers.
• Author names and dates on comments and annotations, and attached files.
• JavaScript, and metadata inside JPEG 2000 images.
• Encrypted PDFs. Remove the password first if you are entitled to.

How the cleaning works

A PDF is a collection of numbered objects: pages, fonts, images, content streams and a catalogue that ties them together, plus a trailer that points at the catalogue. The cleaner reads the file, deletes the descriptive entries (the Info dictionary, XMP streams, private application data and the trailer ID), then walks the document from its catalogue and keeps only the objects it can reach. Everything else is dropped, and the survivors are written out as a fresh file.

Page content, fonts and images are copied as they are. Nothing is rendered, rasterised or re-compressed, which is why the pages look identical and text stays selectable and searchable. Embedded JPEG photos have their metadata segments cut out with the same lossless method as our image tool. Because the file is rewritten, its size and checksum change even when there was little metadata, and that is expected.

Clean a PDF in three steps

1. Drop the PDF anywhere on the page. You can add several at once.
2. Press Clean. The report lists every field it found, with the author, producer and dates highlighted, and flags hidden items such as earlier versions and metadata inside embedded photos.
3. Download. Save the clean file or all of them as a .zip. A file is marked verified clean when a second scan of the output finds none of those items left.

What the report looks like

For a typical file exported from a word processor, the list before cleaning might look like this. The values are examples.

Check the result yourself

• In a PDF reader, open File, Properties. The Description tab shows title, author and dates, and most readers have an XMP view too.
• In a text editor, open the clean file and search for /Author and xmpmeta. PDFs are mostly plain text at the structure level, and our output does not hide objects in compressed streams, so a search is a fair check.
• With a command-line tool, ExifTool lists everything it can find.

exiftool -a -G1 -s clean.pdf

Your document stays on your device

The PDF is opened, rewritten and saved by code running in your browser tab, using the open-source pdf-lib library. Nothing is sent to our servers, which you can confirm in your browser’s network panel while you clean a file. That matters for the documents people most often need to clean: contracts, medical letters, HR files and unreleased reports. Guests can clean files up to 25 MB, free accounts up to 25 MB, Starter up to 100 MB, Pro up to 500 MB and Business up to 1,000 MB.

Fix it at the source as well

Cleaning the finished PDF always works, and you can also stop the metadata being written in the first place.

• Microsoft Word: File, Info, Check for Issues, Inspect Document removes personal properties. When saving as PDF, the Options dialog has a “Document properties” checkbox you can clear.
• LibreOffice: under Tools, Options, Security, Options, tick “Remove personal information on saving”.
• Acrobat Pro: the Sanitize Document command removes hidden information in one pass.
• Everything else: export, then clean the result here.

Many PDFs, or every week

Guests clean 50 PDFs or images a day, one file at a time; free accounts clean 500 a day, 5 files at a time; Starter 25, Pro 100 and Business 500 at a time, with no daily limit. For a whole folder, the API returns a clean PDF and a report for each call, and the MCP server lets an assistant such as Claude or Cursor clean the PDFs in a folder on your computer when you ask. The guide to cleaning many files at once explains which route suits which job. If you also share photos, the image metadata remover works the same way, and the AI metadata cleaner covers generation records and Content Credentials.

Frequently asked questions