About
RedactMy.pics is a small browser tool for hiding private details in a screenshot or photo before you share it.
What it does
You open an image, draw boxes over anything you want gone, and download a PNG. Boxes are filled with a solid color by default, which replaces the pixels underneath instead of smudging them. The export is rebuilt on a canvas, so metadata such as EXIF is not carried over.
Where it falls short
Auto-detect looks for text patterns: email addresses, phone numbers, Social Security numbers, card numbers, and API keys. It reads what the text recognition step manages to see, which means small fonts, low contrast, unusual layouts, and handwriting can all cause a miss.
Faces, addresses, names, and license plates are not detected at all in this version. Those need a box drawn by hand.
Because of that, no detection result should be trusted on its own. Open the downloaded file and look at it before you send it anywhere.
Why it runs locally
Redacting a private document on someone else's server means trusting that server. This tool avoids the question by never uploading the image in the first place. Everything happens in the tab you already have open.
That choice costs something. There is no server to do the heavy work, so the text recognition that powers auto-detect has to run in your browser and is slower than it would be on a server. There is also no way to offer features that need persistence, such as a history or an account.
What it is not
It is not a PDF editor. It works on images, and a PDF needs to be turned into an image first — otherwise the redaction sits on top of a readable text layer. It is not a face blurring tool. It does not remove watermarks, and it should not be used to obscure a copyright notice.
It is also not a guarantee. Redaction removes the regions you select. Whether the result identifies a person depends on everything else in the picture, and that is a judgement the tool cannot make for you.
Verifying the privacy claim yourself
You do not have to take any of this on trust. Open the browser developer tools, switch to the network panel, and load an image. The requests you see are the page itself. There is no request carrying image data, because there is no server code that could receive one.
The audit behind that statement, including the one third-party request the recognition library makes and why it carries no image, is written up on the security audit page. The detection rules and their known failures are on the methodology page.
Contact
You can reach us at jyao10086@gmail.com.