How it works
- 1
Upload your document image
Drag and drop or click to upload a PNG, JPG, or WebP image of a document — a scanned contract, receipt, screenshot, or any image containing text.
- 2
Run local OCR and auto-detect PII
Click 'Analyze Document'. Tesseract.js runs entirely in your browser via WebAssembly, recognizes text, and automatically highlights emails, phone numbers, SSNs, credit cards, and other sensitive patterns.
- 3
Review and export
All detected PII is pre-selected for redaction. Click any region on the image to toggle it. When ready, download a PNG with solid black boxes over all redacted areas.
100% private. OCR runs locally in your browser using WebAssembly — your image never leaves your device.
Drop image here
or click to browse
PNG, JPG, WebP, BMP
Image preview appears here
🛡️ Verify zero uploads — open DevTools → Network tab
Open your browser's DevTools (F12), go to the Network tab, and use this tool. You will see zero outbound requests — all processing runs inside your browser sandbox via WebAssembly or pure JavaScript. Nothing you paste or upload is ever sent anywhere.
Use cases
GDPR / HIPAA compliance
Redact patient names, contact details, and identifiers from medical scans, insurance documents, or HR records before sharing with third parties.
Legal and financial documents
Remove SSNs, account numbers, and personal addresses from contracts, tax forms, or bank statements before filing or forwarding.
Privacy-safe screenshots
Quickly redact emails or phone numbers from screenshots you want to post in bug reports, documentation, or public forums.
Frequently Asked Questions
Does the image get uploaded to a server?
No. All OCR processing happens locally in your browser using Tesseract.js via WebAssembly. Your document image never leaves your device. The only external request is downloading the Tesseract language data (~8 MB) from a CDN on first use — this is cached locally afterward.
What PII types does it detect automatically?
Email addresses, phone numbers (US/international formats), Social Security Numbers, credit card numbers (16-digit patterns), IP addresses, dates (for date-of-birth), URLs, passport numbers, and ZIP codes. Detection is regex-based — it works best on clearly printed text with high OCR confidence.
What image formats are supported?
PNG, JPEG, WebP, and BMP. For best OCR accuracy, use high-resolution images (300 DPI or above) with high contrast between text and background. Handwritten text is supported but accuracy is lower.
Can I redact non-PII regions manually?
Yes. Click any text region on the image to add it to the redaction set — it doesn't have to be auto-detected as PII. This is useful for names, addresses, or any custom sensitive content.
What does the exported file look like?
The export is a PNG file at the original image resolution with solid black rectangles replacing all redacted text regions. The rest of the document is unchanged. The file is named 'redacted.png' by default.
Want unlimited access + saved history?
Pro is $9/month · 30-day money-back guarantee.