Features & how it works

Everything you can do

One tool covers the whole job, from a single portrait to a folder of event photos or a full video clip. Every effect and area below is free for photos, with no watermark and no sign-up.

How the face blurring works

When you add a photo, a face-detection model loads once into your browser and scans the image for faces. For each face it returns a dense mesh of 468 landmarks, points around the eyes, nose, mouth and jaw, so the tool knows exactly where the features are rather than just drawing a rough box. That is what lets you cover only the eyes, only the mouth, or the whole face, and why the blur follows the shape of the head instead of a blunt rectangle.

The chosen effect is then painted onto an HTML canvas: a Gaussian blur for a soft anonymize, a downscaled block grid for pixelation and mosaic, a solid fill for the classic black or white bar, or an emoji drawn over the region. A soft oval mask feathers the edges so the result looks deliberate, not glitchy. Nothing about this step needs a server, the pixels are processed on your own machine and the finished image is handed straight back to you as a download.

Effects and areas in detail

There are ten ways to hide a face. Blur and strong or light blur soften the features while keeping the scene readable. Pixelate and mosaic replace the region with blocks, the look most people associate with censoring. A black or white bar gives the editorial, redacted style often used over the eyes. Noise and invert distort the region for a harsher effect, and emoji stickers drop a playful face on top when you want something lighter than a blur. Each one can target eyes and mouth together, eyes only, mouth only, or the whole face, and the strength slider takes any of them from a subtle veil to a heavy block.

Photos, video and live camera

The same detection runs across three modes. In photo mode you can drop one image or several at once and download them together. In live camera mode your webcam is anonymized in real time and you can save any frame, useful for a quick avatar or a redacted screenshot. In video mode every face is tracked frame by frame and the anonymized clip is exported. Photos and live camera are free and unlimited; video export renders every frame, so it is part of Premium.

Private by design

Because detection and rendering both happen locally in WebAssembly, your images and video never leave your device and are never uploaded, stored or seen by us. There is no account for the free tool and nothing to delete afterwards. That makes it a strong fit for anonymizing photos before you share them under GDPR and similar privacy rules, whether you are a journalist protecting a source, a teacher posting a class photo, or a business publishing an event gallery. (This is not legal advice.)