Blur Faces in Photos

How to use it

1. Open your photo and wait while faces are detected.
2. Click to add a missed area or to uncover one.
3. Set the strength and download.

Good to know

Detection works best on clear, forward-facing faces and can miss small or turned ones. Always look over the result before publishing.

The area is rebuilt from a handful of pixels, so the detail is removed from the saved file.

What gets detected

Detection uses Google's MediaPipe face model, which runs on your device. That model is built for faces reasonably close to the camera, so people far away in a crowd are often missed. Click those spots to cover them by hand. A new area matches the size of the first face found.

Check these before publishing

• Faces in the background, in profile or partly hidden.
• Reflections in windows, mirrors and glasses.
• Name badges, school uniforms and number plates.
• Tattoos or other marks that could identify someone.

How strong is strong enough?

The covered area is reduced to a few pixels and scaled back up, so the detail really is gone from the saved file. Strength runs from 1 to 10. At 8 or more, very little of the face remains. Lower settings can leave enough shape for someone who knows the person to recognise them. The result is a JPG at 93% quality.

Questions