Remove metadata from images
See the GPS coordinates, camera serial number, timestamps and editing history your image is carrying, then remove them without re-compressing the picture.
The tool
What happens to the file
It is read into memory in this tab and parsed in a worker. Removal rewrites the container and copies the encoded image across byte for byte, so the picture is unchanged. Nothing is uploaded, and there is no server to upload it to.
Formats
- JPEGread and remove
- PNGread and remove
- WebPread and remove
- SVGread and remove
- PDFnot supported: removal could not be proven complete
Supported formats
A format is only listed as removable when a round-trip test proves it: strip the file, re-read it, and confirm both that no metadata remains and that the encoded image data is byte-identical. Anything that cannot pass that is listed as unsupported rather than shipped with a caveat.
| Format | Supported | Notes |
|---|---|---|
| JPEG | Read and remove | APP1 (EXIF, XMP), APP13 (IPTC), APP11 (C2PA) and comments. Scan data copied verbatim. |
| PNG | Read and remove | tEXt, zTXt, iTXt, eXIf, tIME and caBX. Transparency, colour and animation chunks kept. |
| WebP | Read and remove | EXIF, XMP, ICCP and C2PA chunks, with the VP8X flags and RIFF size corrected. |
| SVG | Read and remove | Editor metadata, RDF, titles and comments. Re-serialised, so geometry is preserved but formatting may change. |
| Not yet | Removal is not offered because it cannot currently be proven complete: metadata also hides in annotations, form fields and earlier revisions. |
How removal actually works
Image files are containers. A JPEG is a stream of marker segments, a PNG is a sequence of chunks, a WebP is RIFF chunks. The picture lives in some of them and the metadata in others, so removal means writing out a new file containing only the segments worth keeping and copying the image payload across untouched.
The alternative is drawing the image to a canvas and exporting it, which is what most in-browser tools do. It is worse in every way that matters: it recompresses the picture, loses the colour profile, and produces a different file on every browser. It is not used here.
Common questions
- Does removing metadata reduce image quality?
- No. For JPEG, PNG and WebP the file is rebuilt without the metadata containers while the compressed image data is copied across byte for byte. The picture is never decoded and re-encoded, so the result is pixel-identical to the original. This is tested by comparing the encoded payload before and after.
- Why is the orientation flag kept by default?
- Because removing it visibly rotates photographs. Many cameras store the image sideways and record an orientation tag that viewers use to rotate it back. Strip the tag and the photo appears rotated. It is a separate opt-in for that reason, and it is the single most common bug in metadata removers.
- Why is the colour profile kept by default?
- An ICC profile tells software how to interpret the colour values. Removing it from a wide-gamut image shifts the colours, usually towards oversaturation. It is offered as a separate option with that warning attached.
- What does it find that I might care about?
- GPS coordinates precise to a few metres, the camera or phone model, sometimes a unique body serial number, the exact capture time, the name and contact details in authorship fields, and which software edited the file. These are grouped by what they mean rather than by tag name.
- Are my images uploaded?
- No. Parsing runs in a worker inside your browser tab and there is no server, so your image is never transmitted. This page carries Google advertising, which contacts Google and never your file. If you would rather work with no third-party code in the page at all, the same tool is at /app/metadata, where a second content security policy blocks every outbound request.