Automatically finds emails, phone numbers, SSNs, Aadhaar, PAN, and card numbers. Redacted pages are rebuilt as images so the text is genuinely gone — not just covered.
No uploadAuto-detects PIITrue removal, not a cover-upFree forever
Drop your PDF here
Click anywhere to browse
Browse PDF
Files stay on your device · No signup
—
—
Scan for
Scanning pages…
Page 1 of 1
Matches on this page —
ℹOnly pages with a confirmed match are rebuilt as images. Pages with nothing redacted stay as the original, fully searchable vector content.
No upload Processed in browser True removal
Redact a PDF — real removal, not a black box on top
Most "redaction" happens by drawing an opaque shape over sensitive text — but the original text is often still there underneath, extractable by copying the text or opening the PDF's raw content. This is a well-documented failure mode behind several public redaction leaks over the years.
This tool avoids that failure by construction: any page with a confirmed match is rebuilt from scratch as a flattened image, with the redaction boxes baked directly into the pixels before the page is reassembled. There's no text layer left on that page to extract — the data is gone, not hidden. Pages with nothing to redact are left completely untouched, still vector, still searchable.
Detection is a starting point, not a guarantee. Regex pattern matching can miss unusual formats and can occasionally flag something that isn't actually sensitive. Always review the highlighted matches on each page before downloading, and use the manual draw tool for anything the patterns don't catch — including scanned pages, which have no text layer for automatic detection to work with.
Want to also strip embedded scripts, attachments, or flatten form fields? See Flatten PDF. Need to clean up metadata on its own? Use the PDF Metadata Editor.
Frequently asked questions
It really removes it. Any page with a confirmed redaction is rebuilt from scratch as a flattened image with the black box baked into the pixels — there is no text layer left to copy, search, or extract on that page. This avoids the classic "redaction fail" where black boxes are just a cosmetic layer sitting on top of selectable text underneath.
A page that gets redacted loses text selection and search on that specific page, since it becomes an image. Pages with no confirmed matches are left completely untouched — still vector, still searchable, at full quality.
Email addresses, phone numbers, US Social Security Numbers, Indian Aadhaar numbers, PAN numbers, and credit card numbers (validated with a checksum to reduce false matches). You can enable or disable each category, and draw additional redaction boxes manually for anything the patterns miss.
No — it's pattern matching, not verified identification. Always review the highlighted matches before downloading. Uncheck false positives, and manually draw boxes over anything sensitive that wasn't caught automatically.
Automatic detection needs a text layer, so it won't find anything on a pure image scan. You can still redact scanned pages by drawing boxes manually over the sensitive areas in the preview.
No. Detection and redaction both happen entirely in your browser using pdf.js and pdf-lib. Your file never leaves your device.