Client-Side PDF Redaction & Sanitizer Tool
Permanently destroy sensitive text in PDFs by burning black redaction bars into rasterized pages (scale = DPI / 72), rebuilding a zero-text-layer PDF, and re-parsing the output with PDF.js to prove zero text residue.
- Nothing is uploaded — turn on airplane mode after loading to verify 100% local sanitization.
- No account, no email
- No data leaves your browser
1. Document & rasterization DPI
confidential-kyc-dossier.pdf (sample)
2. Auto-detect PII regex patterns & custom strings
Toggle automatic pattern redaction and verifier targets
Why annotation black bars are unsafe — and how true redaction works
Why standard PDF black boxes fail: Drawing a black rectangle in a typical PDF editor only appends a vector path or /Annot overlay on top of the original /Contents stream. The underlying characters remain intact in the file—anyone can press Ctrl+A / Ctrl+C or delete the vector box to expose SSNs, names, and bank numbers.
How this tool guarantees irrecoverability: Each page is rendered via PDF.js onto an HTML5 Canvas at scale = DPI / 72 (2.08×). Solid #000000 pixels are painted directly over every redaction box, overwriting the underlying RGB bytes. Finally, pdf-lib builds a brand-new PDF containing only flat raster images—and our built-in Post-redaction verifier re-parses the generated PDF bytes with pdfjs.getDocument() to confirm that 0 sensitive strings can be searched.
Result
True pixel-level sanitization @ 150 DPI
pdfjs re-parse of the generated PDF bytes to verify that marked strings cannot be searched.Included in Pro. Unlock once, keep it on this device.
Interactive redaction canvas
Click and drag on the document to draw custom black boxes
| Page | Rule / Source | Normalized Box (x, y, w, h) | Action |
|---|---|---|---|
| P1 | SSN: 123-45-6789 | (31%, 23%, 24%, 3%) | |
| P1 | IBAN: DE89 3704 0044 0532 0130 00 | (31%, 28%, 48%, 3%) | |
| P1 | Email: [email protected] | (31%, 34%, 42%, 3%) | |
| P1 | Phone: +1 (415) 555-0198 | (31%, 40%, 31%, 3%) | |
| P1 | Card: 4532 0151 1283 0366 | (31%, 46%, 38%, 3%) | |
| Total: 5 | Sample document | Text layer: Stripped |
What people hate about the other tools
We read the reviews of the incumbents before writing this page. Every complaint below maps to a decision in the tool above.
Cover-up boxes in a normal editor often leave the text underneath selectable. Before I send anything with account numbers or SSN, I export and paste-check.
In this toolDestroys underlying text streams via pixel sanitization and proves zero residue in a live Paste-Check view.Uploading tax returns, legal contracts, or HR records to third-party web redactors means trusting their servers; turn on airplane mode to test if it's local.
In this toolRuns 100% in your browser with a live 0-bytes-uploaded badge and full airplane-mode offline support.Distrusting online tools forces users to print documents and re-scan them or convert every page to heavy images, ruining clarity and bloating file size.
In this toolSanitizes redacted regions and strips hidden OCR/metadata cleanly in-browser without physical printing.
One-line difference: 100% offline-capable true PDF sanitization that destroys underlying text/OCR streams with a built-in Paste-Check verifier.
Free forever, Pro when you need the paperwork
Free no signup, no card
- Unlimited manual blackout with destructive pixel/page flattening
- Complete metadata, XMP, form field & hidden OCR text layer removal
- Built-in Paste-Check raw text & Ctrl+F residue proof verifier
Pro $19 once
- Multi-file batch redaction pipeline entirely in browser memory
- Auto-sweep regex scanner for SSNs, emails, phones & account numbers
- Selective vector sanitization keeping unredacted text sharp
- Exportable local Redaction Audit Certificate for compliance records
Pricing benchmarked against what the incumbents charge. See the full monetization map → · Already paid? Restore access
The tool runs locally in your browser — nothing you type is uploaded. Ad slots on this page are served by Google. See all 27 tools · Privacy · How these are paid for