How it works
- 1
Open your PDF. The Redact tool is already selected.
- 2
Drag boxes over what to hide, or type a word, name or number and mark every match.
- 3
Select Download. The content under the boxes is removed from the saved copy.
Uploading to redact defeats the point
You redact because a document holds something others must not see: a patient's name, a national ID number, a client in a legal filing, salary figures. Sending the unredacted original to a server so it can be blacked out means the sensitive version has already left your hands.
Here, the redaction is done by the PDFium engine compiled to WebAssembly, in a background worker inside your tab. It works on a copy in memory, removes what is under each box and writes a new PDF that you download. The original stays untouched on your disk, and the unredacted copy is gone from memory when you close the tab.
What gets removed, not just covered
A black rectangle drawn on top is not redaction: the text underneath can still be copied out. When you download, this tool deletes the content itself.
- Text under a box is deleted from the page, so it cannot be selected, searched or copied.
- Simple shapes fully inside a box are deleted. Comments, links and form fields under a box are removed, and typed form values are wiped.
- If a box covers a photo, a scan or complex drawing (which might be text drawn as outlines), that page is saved as an image at 200 DPI so nothing underneath can be recovered. You are told which pages this happened to.
- Words you redacted are also replaced in the document title, author and other properties, and in bookmark titles.
Check it yourself
Two checks. First, privacy: open DevTools (F12), Network tab, then redact and download. No outgoing request matches your file's size; you may see the tool's own code and background sign-in checks that every page makes. Redaction does not fetch fonts or anything else. You can also open the file, turn off Wi-Fi and redact and download offline.
Second, the result: open the downloaded PDF, press Ctrl+A (Cmd+A on Mac) on a redacted page and copy. The removed words are not in what you paste. Searching for them finds nothing.
Limits to know
Search finds words only in PDFs with real text. A scanned page has no text to search, and there is no OCR, so on scans draw the boxes by hand. Covering a scan with a box turns that page into an image, which removes what is under the box.
Pages saved as images lose selectable text on that page. Very large scanned documents use more memory on phones. Redaction is free, and the saved file is not watermarked.
Private by design: this tool runs in your browser with WebAssembly. Your file is never uploaded, and it is gone from memory when you close the tab.