Neura Market
HideMyData – macOS PII Removal with Local AI and OCR logo

HideMyData – macOS PII Removal with Local AI and OCR

Free
FreeFree tier
Type
Open Source
Company
mkbula
Visit Website GitHub (1 stars)

About HideMyData – macOS PII Removal with Local AI and OCR

HideMyData is a native macOS application for local, AI-powered removal of personally identifiable information (PII) from PDFs and images. It runs entirely on-device using Apple Silicon, ensuring no data ever leaves the machine. The tool combines Apple Vision OCR (for scanned documents and embedded fonts), an OpenAI-based privacy filter (MLX 8-bit quantized) to detect names, emails, phones, addresses, dates, and IDs in context, plus manually maintained regex patterns for common identifiers like IBAN, SSN, MAC, IP addresses, JWT, API keys, and crypto wallets. Users can choose between solid black or frosted glass blur redaction styles, manually add or remove redaction rectangles before saving, and the resulting PDFs are permanently rasterized so original text and glyphs are completely removed. Built with Swift 6, SwiftUI, MLX-Swift, Apple Vision, PDFKit, and OpenMedKit, and released under GPL-3.0.

Key Features

Fully on-device processing — model runs locally, nothing leaves your machine
Supports PDF and image input with shared detection and redaction pipeline
Apple Vision OCR handles scanned PDFs, images, and PDFs with embedded fonts that hide text from selection
AI detection via OpenAI privacy-filter (MLX 8-bit) catches names, emails, phones, addresses, dates, IDs in context
Manually maintained regex for IBAN, SSN, MAC, IPv4/v6, JWT, API keys, crypto wallets and more
Two redaction styles: solid black or frosted glass blur
Manual editing — add or remove redaction rectangles before saving
Permanent on save — pages are rasterized and rebuilt, original text and glyphs are gone
Built with Swift 6, SwiftUI, MLX-Swift, Apple Vision, PDFKit, OpenMedKit
Free and open source (GPL-3.0)

Pros & Cons

Pros
  • Fully local and private — no data sent to external servers
  • Combines AI detection with regex for comprehensive PII coverage
  • Works with both PDFs and images, including scanned documents via OCR
  • Permanent redaction ensures information cannot be recovered
  • Open source and free with no usage limits
  • Two redaction styles for flexibility
  • Manual editing gives users fine control over what is redacted
Cons
  • macOS only (requires macOS 26 or later) — no Windows or Linux support
  • Requires Apple Silicon (MLX backend does not run on Intel)
  • First launch requires downloading a ~1.5 GB model from Hugging Face
  • Relatively new project with limited community and fewer releases
  • No cloud or cross-platform sync option
  • Only outputs redacted PDFs (no other output formats mentioned)

Best For

Redacting personal information from PDF documents before sharingRemoving sensitive data from scanned images and documentsAnonymizing legal, medical, or financial records for research or reviewCleaning up screenshots and image files that contain PIIProtecting privacy when sending documents to third parties

FAQ

Is HideMyData free?
Yes, HideMyData is completely free and open source under the GPL-3.0 license.
What are the system requirements?
HideMyData requires macOS 26 or later and an Apple Silicon Mac (the MLX backend does not run on Intel).
Does HideMyData send my data to the cloud?
No. The model runs entirely on-device; nothing ever leaves your machine.
What types of PII does it detect?
It uses an AI privacy filter to catch names, emails, phone numbers, addresses, dates, and IDs, plus regex for IBAN, SSN, MAC, IPv4/v6, JWT, API keys, and crypto wallets.
How do I install HideMyData?
Download the latest .dmg from the GitHub Releases page or build from source using Xcode.