Settings

What is this photo carrying?

Most photos from a phone or camera embed far more than the picture. Drop one in to see it.
No photo loaded yet.

Redact and export

Drag on the image to cover anything that identifies a person or place. Redaction replaces the pixels themselves, so there is no hidden original layer underneath: unlike a black box drawn over a PDF.
Original
Cleaned idle
Drag on this image to redact a region.
The exported file is re-encoded from raw pixels, which drops every metadata block: EXIF, GPS, XMP, IPTC, maker notes and the embedded thumbnail. Verify it yourself: load the downloaded file back into this page and the report above should come up empty.

What this does and does not do

It does: read and remove embedded metadata, including GPS coordinates, camera make, model and body serial number, capture timestamps and software fingerprints. It permanently destroys pixels in the regions you mark, so a redacted face or document cannot be recovered from the exported file.

It does not: claim to defeat any specific face-recognition model. Producing adversarial examples against a system like ArcFace or FaceNet requires that model's own weights and gradients; noise generated without them does not reliably fool it, and we would rather tell you that than sell you a placebo. If you need a face not to be matched, redact it: removing the pixels works against every matcher because there is nothing left to match.

Also worth knowing: stripping metadata does not anonymise what the photo shows. Street signs, reflections, window views and screen contents identify a place perfectly well on their own.

Just want to check a photo without unlocking anything? Use the free photo metadata checker.