ScreenRedact

ScreenRedact is a focused, no-frills Mac utility that does one thing and does it well: redact text from images using on-device machine learning. Developed by Ahmad Tibi, this free app is designed for users who need to quickly sanitize screenshots, scanned documents, or any image containing sensitive information before sharing. Its core value proposition lies in privacy: because all processing happens locally using Apple’s Vision framework, your images never leave your Mac. This makes it an attractive choice for developers, IT professionals, and privacy-conscious individuals who want a lightweight alternative to cloud-based redaction tools.

Where ScreenRedact stands out is its simplicity. After opening an image, the app automatically detects text regions and highlights them with clickable bounding boxes. A single click redacts a box, turning the text into a solid black rectangle. The interface is minimal, with no layers, undo history, or fine-grained controls. This is both a strength and a limitation. For users who need to redact a handful of items from a screenshot before posting it on Slack or GitHub, the workflow is nearly instantaneous. The app also supports copying the redacted image directly to the clipboard, which integrates smoothly into fast-paced sharing workflows.

However, the tool’s simplicity comes with notable constraints. ScreenRedact only saves images as PNG files, which can be a problem for users who need JPEG compression for smaller file sizes or other formats for specific workflows. There is no batch processing, so redacting multiple images requires manual repetition. The app requires macOS 14.0 (Sonoma) or later, excluding older Macs. Additionally, the text detection, while generally accurate for clean screenshots, can struggle with complex layouts, small fonts, or non-standard typefaces. In our tests, it occasionally missed text in images with heavy backgrounds or overlapping elements, requiring manual adjustment that the app does not support. Users expecting pixel-perfect redaction may need to double-check results.

The ideal user for ScreenRedact is someone who values speed and privacy over advanced editing features. Developers sharing code snippets can quickly hide API keys or passwords. IT professionals redacting internal documentation—such as network diagrams with IP addresses or employee names—will appreciate the offline processing. Privacy advocates will find the on-device ML approach aligns with zero-trust principles, though power users may miss features like blurring instead of blacking out, or the ability to redact non-text elements. For non-technical users, the app’s simplicity is a double-edged sword: it’s easy to use but offers no guidance on best practices, such as ensuring redactions are irreversible (black rectangles can sometimes be lifted if not opaque enough).

In practice, ScreenRedact fills a narrow but genuine niche. It’s not a replacement for full-featured image editors or enterprise redaction tools, but for quick, private redaction of text in images on a modern Mac, it gets the job done. Buyers should evaluate whether the PNG-only output and lack of batch processing fit their workflow, and whether their macOS version is supported. If those constraints are acceptable, ScreenRedact offers a free, efficient, and privacy-respecting solution for a common but often overlooked task.