Free tool
Blur faces in a photo
Add a photo, and the tool finds the faces and blurs them. You can move, resize or remove each box, and draw new ones over faces, names or plates. The saved photo has the pixels replaced, not covered, and no metadata. The photo is not uploaded. The work happens in this browser.
Runs on this device. Your photos are not uploaded.
Drop a photo here
JPEG, PNG, WebP or HEIC (HEIC in Safari). One photo at a time. It stays on this device.
JPEG, PNG, WebP or HEIC (HEIC in Safari). One photo at a time. It stays on this device.
Drag on the photo to draw a box. Drag a box to move it, or its corner to resize it. With the keyboard: Tab to a box, arrow keys move it, Shift and arrow keys resize it, Delete removes it.
How it works
- Find. The tool runs BlazeFace, a face detection model that Google publishes for phones and browsers. It checks the whole photo, then smaller parts of it, so it also finds small faces in a group photo. It checks each part again as a mirror image, which finds faces that one pass misses.
- Check. Each face gets a box that covers the head. A dashed box means the tool is less sure, so look at it. Remove boxes you do not want, and draw boxes over any face, name tag or number plate it missed.
- Replace. Blur turns each box into about 64 averaged colors and smooths them, with a soft edge. Solid fill paints the box black. Both change the pixels themselves.
- Save. The tool draws a new photo with the boxes applied and saves it. The new file has no metadata. Your original photo does not change.
The face detection files load from yorkvault.com the first time you add a photo, about 13 MB, and your browser keeps them. No part of the photo goes anywhere.
Questions people ask
Does this tool upload my photo?
No. The photo stays in your browser. The face detection model and its code come from this site once, then run on your device. To check, load this page, add a photo once so the model downloads, then turn off Wi-Fi and add another photo. Face detection still works.
Can someone remove the blur?
Not from the saved file. The tool replaces the pixels in each box with about 64 averaged colors, then smooths them. The face details are not in the new file, so no app can sharpen them back. Solid fill replaces the pixels with black. An editing app that only lays a blur on top can be undone. This tool does not work that way.
How do I blur faces in a photo on iPhone?
Open this page in Safari on your iPhone, tap Choose a photo, and pick the photo. Check the boxes, then tap Download photo. Safari saves the new photo in the Downloads folder in the Files app. From there, you can share it or save it to Photos.
Why did the tool miss a face?
Face detection finds most faces that look at the camera. It can miss faces that are small, turned to the side, in shadow, behind glasses and hair, or partly covered. Drag on the photo to draw a box over any face it missed. Always check every face before you share.
Does the blurred photo keep its location and camera data?
No. The tool makes a new image file, which has no EXIF, XMP or IPTC data. It holds no location, date, camera model or serial number. Your original photo does not change.
Sources
- Convert image file types using Preview on Mac (Apple) · checked Oct 3, 2026
- Face detector guide: BlazeFace models (Google MediaPipe) · checked Oct 3, 2026
- BlazeFace: Sub-millisecond Neural Face Detection on Mobile GPUs (arXiv) · checked Oct 3, 2026
York Vault
Blur before you share, inside York Vault
York Vault keeps your photos encrypted on your phone. Its photo editor finds faces and blurs or blacks them out in a new copy, and you can box any text by hand. When you share the copy, York Vault removes the location and camera details by default.
