TL;DR: Upload a PDF and let AI find and redact emails, phone numbers, SSNs, names, etc. — all detection happens in your browser via BERT-NER.
Auto-redact PII from PDF online for free — local BERT-NER detects names, emails, phone numbers, addresses, and SSNs in your browser before the backend permanently applies approved redactions.
Reach for Smart Redact PDF (AI) when you want the job done without handing your file to a data broker. Browser-native tools stay on your device; the rest use a single-request isolated container that deletes the file right after responding. MIT-licensed and ad-free.
Auto-redact PII from PDF online for free — local BERT-NER detects names, emails, phone numbers, addresses, and SSNs in your browser before the backend permanently applies approved redactions. Use it when you need a quick, private, no-account way to handle a pdf in the browser, or when you want an auditable open-source alternative to iLovePDF, Smallpdf, Adobe Acrobat, PDF24, and Sejda. The page at /tool/smart-redact is designed for one clear job: upload or provide the input, choose only the options that matter, and download the result without creating an account or passing through a sales funnel.
PDF operations that need server-side libraries run inside the PrivaTools container and return a fresh download; browser-only PDF helpers stay on-device. Temporary input and output files are not used for analytics, model training, advertising profiles, or product telemetry. The public demo uses anonymous page-view analytics only; file bytes, extracted text, filenames, passwords, signatures, and generated results are outside that analytics path. If your organization needs stricter controls, you can self-host all 214 PrivaTools utilities and keep processing on your own infrastructure.
Before running Smart Redact PDF (AI), confirm that the source file opens correctly and that you have permission to process it. Keep an untouched original, run one operation at a time when quality matters, and use Pipeline when you want repeatable multi-step output. For sensitive material, review the downloaded result before sharing it. For large files, give the browser time to finish the download and avoid refreshing the page mid-run. If a password, damaged upload, unsupported codec, or malformed document blocks processing, PrivaTools returns a plain-language error so you can pick the next recovery step instead of guessing.
Smart Redact PDF (AI) is intentionally narrow: it does one pdf task and hands the result back as a normal download. That makes the output easy to inspect, rename, archive, attach to email, or feed into another tool. If you need a repeatable workflow, save the page, bookmark a Pipeline recipe, or self-host the API so the same steps can run from internal scripts. The interface avoids accounts and cloud folders on purpose: the safest default for private files is to process only the current request, return the result, and leave long-term storage under your control.
Smart Redact PDF (AI) runs in any modern browser on Windows, macOS, Linux, Android, and iOS — there is nothing to install, no extension to add, and no desktop app to keep updated. Because the interface is a single page, you can bookmark it, send the link to a colleague, or open it on a phone and get the same result you would on a laptop. There are no watermarks stamped onto your output, no sign-in wall before the download, and no paid tier that unlocks the “real” version later — the Smart Redact PDF (AI) you see is the complete tool. For teams that would rather keep everything in-house, the same endpoint ships in the MIT-licensed, self-hostable build, so you can run it behind your own firewall with identical behaviour and no outbound calls. That combination — instant in the browser for individuals, fully self-hostable for organizations — is what keeps a private file genuinely private from upload to download.
Names (persons and organizations), emails, phone numbers, postal addresses, SSNs and other government IDs, credit card numbers, dates, locations, and IP addresses. The BERT-base-NER model also catches custom patterns via regex hooks — useful for proprietary identifiers like case numbers or patient IDs.
Entity detection runs in your browser using @huggingface/transformers WebAssembly. The PDF text never leaves your machine during scanning. When you click Apply, the PDF and selected strings are sent to the isolated backend so PyMuPDF can permanently bake in the redactions; temporary files are unlinked after the response.
No. The backend applies real PyMuPDF redactions which permanently destroy the underlying glyphs and image pixels under each redaction rectangle. The redacted file cannot be 'unredacted' even with forensic tools — there is nothing left to recover.
Regular Redact PDF requires you to draw rectangles manually over each piece of sensitive content. Smart Redact auto-detects PII entities so you only need to review the model's suggestions and click 'apply' on the ones to redact. For known-pattern data (SSNs, credit cards) accuracy is near-perfect; for free-text names, you should review.
Yes — and arguably safer than manual redaction because the model catches PII you'd miss. Detection runs in-browser (the document text never uploads); only the final redacted file is sent to the server for the permanent-redaction pass, then unlinked immediately. The whole pipeline is open source for verification.
Detection runs in-browser so it's bounded by browser memory (typically 500 MB practical). The server-side redaction pass accepts up to 500 MB. No daily or monthly quota.
The default BERT-base-NER model is English-only with strong coverage. Multilingual NER models are on the roadmap. For non-English documents, you can use Smart Redact for the pattern-based entities (numbers, emails, dates) which work across languages, plus regular Redact PDF for free-text.
No — no account, no email, no sign-up. The tool loads in your browser; privacy guarantees are architectural.
See how PrivaTools compares to iLovePDF, Smallpdf, Adobe Acrobat, and other free PDF tools.