What documents does PrivateRedact support?

The app reads text PDFs, scanned or image PDFs (with OCR), DOCX, TXT, and images. Redacted output can be exported as a secure rasterized PDF or as format-preserved DOCX.

How much does a license cost and how long does it last?

PrivateRedact is free to load, detect, and review documents. A license is required only to export redacted output, purchased one-time from Gumroad with no subscription; you can use it forever once purchased.

Does PrivateRedact work offline?

Yes. All processing runs on your machine with no file, hash, metadata, or content transmitted anywhere. The app installs Ollama (the local AI engine) on first use, and everything operates offline thereafter. License verification calls Gumroad, but the app works offline for a grace period after activation.

What operating systems are supported?

Currently macOS (Apple Silicon) and Windows (x64). Intel/x64 macOS and Linux AppImage targets are on the roadmap.

Large Language ModelsAI Safety & AlignmentOpen-Source AIHacker NewsAug 10, 2026Aug 11, 2026, 01:01 JST3 min read

PrivateRedact: offline PII redaction with local LLM, no cloud upload

Send on LINE

Key takeaway

PrivateRedact is a free-to-try desktop app that redacts sensitive personal information from documents entirely offline using a hybrid regex and local AI model, with no data leaving your machine.
You can load and review redactions for free; only exporting requires a one-time license purchase.
Unlike cloud tools that require uploading the file, or naive tools that black out whole lines or leave recoverable text beneath redactions, PrivateRedact redacts only the sensitive value and rasterizes the PDF so redactions are permanent.

3 Key Points

What happened
Muhammad Karim has released PrivateRedact, a desktop application that redacts personally identifiable information (PII) from documents entirely on your machine using a hybrid regex and local LLM pipeline. The tool supports text PDFs, scanned PDFs with OCR, DOCX, TXT, and images, and exports securely rasterized PDFs or format-preserved Word files with all redactions baked in.
Why it matters
Cloud-based redaction tools require uploading the sensitive document to a server—defeating the privacy goal—while naive desktop tools either force manual redaction or black out entire lines, leaving the underlying text selectable. PrivateRedact solves this by redacting only the sensitive value (e.g., "555-1234") while preserving the label ("Phone:") and surrounding context, and it rasterizes the PDF so redactions cannot be undone or stripped away.
What to watch
Loading, detecting, and reviewing redactions are free; exporting requires a one-time Gumroad license key with no subscription. The app auto-installs Ollama (the local AI engine) on first use and picks model size based on available RAM (~8 GB minimum). macOS (Apple Silicon) and Windows (x64) are currently supported; Intel/x64 Mac and Linux AppImage targets are on the roadmap, as is signed Windows builds (currently unsigned).

Ask the AI about this article →

Summaries like this, in your inbox every morning.

Get news like this every morning — free

Context & Analysis

PrivateRedact addresses a fundamental tension in document privacy: conventional cloud redaction tools solve the technical problem of PII detection by introducing a security problem—the need to upload the sensitive document to a third party. Conversely, existing offline tools either force slow, error-prone manual redaction or apply crude whole-line blackouts that leave the underlying text layer intact and recoverable. The application bridges this gap by combining pattern-based detection (regex for known formats like SSNs, credit card numbers, IBANs, and dates of birth) with semantic detection via a local LLM (for diagnoses, names, addresses, and codenames), then exporting a rasterized PDF where redactions are baked into the image layer itself—making un-redaction or text recovery technically infeasible.

The tool's design choices reflect security-first principles: metadata is stripped by default (PDF Info/XMP, DOCX properties, image EXIF), redacted PDFs can optionally be encrypted, and the app includes a post-export leak scan that re-OCRs the output to confirm no PII survived. The licensing model—free trial for all detection and review work, paid export only—lowers the barrier to testing against real documents, while the one-time purchase (via Gumroad) avoids subscription friction for privacy-conscious users. System requirements are modest (~8 GB RAM minimum), and the app auto-selects the local model size based on available memory, removing a technical barrier for non-expert users.

FAQ

What documents does PrivateRedact support?: The app reads text PDFs, scanned or image PDFs (with OCR), DOCX, TXT, and images. Redacted output can be exported as a secure rasterized PDF or as format-preserved DOCX.
How much does a license cost and how long does it last?: PrivateRedact is free to load, detect, and review documents. A license is required only to export redacted output, purchased one-time from Gumroad with no subscription; you can use it forever once purchased.
Does PrivateRedact work offline?: Yes. All processing runs on your machine with no file, hash, metadata, or content transmitted anywhere. The app installs Ollama (the local AI engine) on first use, and everything operates offline thereafter. License verification calls Gumroad, but the app works offline for a grace period after activation.
What operating systems are supported?: Currently macOS (Apple Silicon) and Windows (x64). Intel/x64 macOS and Linux AppImage targets are on the roadmap.

Hacker NewsRead Original Article