Your photos remember more than you do.
Camera make, GPS coordinates, timestamps, even the software that edited it — devices quietly stamp all of it into every image file. See what's hiding in yours, then strip it clean before you share.
A metadata cleaner removes the hidden data stored inside an image file — information you never see when you look at the picture itself. Most photos carry a quiet paper trail: the camera or phone that took them, the exact GPS coordinates, the date and time, and a record of any software used to edit them.
AI-generated images carry their own version of that trail. Files produced by tools like Stable Diffusion, Midjourney, or Adobe Firefly can embed the prompt, seed, sampler settings, or content-provenance credentials that describe exactly how and where the image was made.
People strip this data for two main reasons. The first is privacy — GPS tags on a photo taken at home can expose your address, and device identifiers can tie separate photos back to the same person. The second is control over provenance — creators often prefer not to publish the generation parameters or authenticity credentials baked into an AI image, the same way they'd remove GPS or camera data from an ordinary photo before sharing it.
This tool handles both cases the same way, in a single pass: everything runs locally in your browser. Your file is never uploaded anywhere, and there's no account required.
Re-encoding an image through a blank canvas carries pixels forward and nothing else — so every chunk-based or tag-based data category below is dropped in the same pass.
Camera make and model, exposure settings, timestamps, and software tags — detected and listed before you clean the file.
Geotagged coordinates embedded by phones and cameras, with a map link so you can see exactly what you'd be exposing.
Creator name, copyright notices, keywords, and caption fields some editors and cameras write into a file.
tEXt, iTXt, and zTXt chunks that Stable Diffusion interfaces like Automatic1111, ComfyUI, and Forge use to store prompts, seeds, and sampler settings.
Content credential manifests embedded by Adobe Firefly, Photoshop, and other Content Authenticity Initiative tools.
Tool and model fingerprints written into a file by DALL·E, Midjourney, and similar image generators.
Small preview images some cameras and editors save inside the file, which can retain detail cropped from the visible photo.
Re-encoding produces a brand-new file, so the exported copy no longer matches the cryptographic hash of the original.
Every tag category above is stripped in the same step — no separate tool needed for GPS data, AI signatures, or provenance credentials.
The file is read and re-encoded using your browser's own APIs. There's no server in between, so there's nothing to intercept, log, or leak.
Core cleaning is free, with no sign-up, watermark, or file limit standing between you and a clean download.
Remove generator signatures from Stable Diffusion, Midjourney, and DALL·E images before sharing them.
Strip embedded metadata and provenance data from AI-generated content before posting to Pinterest, Instagram, or Facebook.
Clean metadata from AI artwork before uploading to Etsy, Amazon Merch, or other POD platforms.
Use AI-generated images in posts and marketing material without embedded data revealing their origin.
Keep AI-assisted content looking natural and professional by removing the traces that give it away.
Remove GPS location data and camera information from ordinary photos before sharing them online.
Four steps, all inside your browser tab.
Loaded straight into your browser via the HTML5 File API. No network request is made and no file leaves your device.
The tool reads embedded EXIF, GPS, and IPTC tags and lists them so you know exactly what's inside the file before you touch it.
Cleaning redraws the image onto a blank canvas and exports it fresh. Only pixel data crosses that boundary — chunk-based metadata, provenance manifests, and generator tags don't survive the redraw.
The clean file downloads immediately, in your original format or converted to JPG, PNG, or WEBP.