Redact a PDF — remove text permanently

Black out names, numbers and paragraphs so they are gone from the file itself, not merely hidden behind a rectangle a reader can lift off.

How to Redact the PDF

1
Drop your PDF in
Choose the file or drag it onto the page. Your browser opens it and draws the first page; nothing leaves the tab at any point.
2
Select the areas or search for the text
Drag a box over anything that has to go, or type a name and let the tool find every occurrence across the document and list them for review.
3
Download the redacted file
The pages you marked are rebuilt as images, the file is checked to confirm the text is gone, and only then does it download.

Covering text is not removing text

A black rectangle drawn over a paragraph changes nothing underneath it. The words stay in the page's content stream, and anyone can select them, copy them, or pull them out with a command-line extractor in about a second. Court filings, police reports and government contracts have leaked this way, over and over. This tool takes the only approach that works in a PDF: every page you touch is redrawn as a raster image at 150, 200 or 300 DPI, the boxes are filled onto those pixels, and the new page carries the picture instead of the original text objects. The words are not covered. They are not in the file at all.

Only the pages you marked change

Rasterising a whole document to hide two lines would inflate the file several times over and destroy text search on pages that had nothing to hide. Untouched pages are copied across byte for byte, so they stay selectable, searchable and small. After the file is assembled it is reopened and read back: if any text is still reported on a redacted page, the result is thrown away and nothing is offered for download.

Why the removed text is really gone

Your PDF is opened, rendered and rewritten inside this browser tab; a document full of personal data has no business travelling to a server. The redacted pages are rebuilt from pixels, so no hidden text layer survives them, and the finished file is re-read to prove it before you ever see a download button.

Frequently asked questions

Why can't I just draw a black rectangle over the text?

Because a rectangle is drawn on top while the text stays in the content stream below it. Select the area in any viewer, copy it, and the supposedly hidden words appear in your clipboard. Documents have leaked this way for years, which is why we rasterise the page instead.

Can the redacted text be recovered from the file?

No. The original page is not copied into the result at all. It is replaced by a flat image of itself with the boxes already burned into the pixels. There is no text object, no font mapping and no hidden layer left to recover anything from.

Why did the redacted pages get bigger?

A page of text is a few kilobytes of instructions; a photograph of that same page at 200 DPI is a few hundred kilobytes of JPEG. That is the price of removal. Lower the resolution to 150 DPI, or run the result through a compressor afterwards.

Why does text search stop working on redacted pages?

Those pages are images now, so there are no words for a viewer to find; only the pages you did not touch stay searchable. Run OCR on the finished file to rebuild a text layer; it will describe what is visible, which no longer includes what you removed.

Are the metadata and the author name removed too?

Yes, by default. The title, author, subject, keywords, creator and producer fields are dropped along with the XMP packet that repeats them. A name hidden on page four is worthless if the same name still sits in the document properties.

Does my file go to a server?

No. Reading the PDF, rendering the pages, filling the boxes and writing the new file all happen in your browser. Nothing is uploaded, queued or stored, and a document you redact here never exists anywhere we could reach it.

Can I search with a regular expression?

Yes. Switch the pattern option on and search for things like a social security number or a date format across the whole document. Every match is listed with its page number, and you can untick any occurrence that should stay visible.

What if my PDF is a scan with no text layer?

Searching needs real text, so on a scan it finds nothing. Select the areas by hand instead; the result is identical, because the page is rasterised either way. If you want to search a scan, run OCR on it first and come back.