Meta-Stripper

Meta-Stripper is a free browser tool that removes the metadata riding along inside image files. It clears EXIF blocks, GPS coordinates, XMP and IPTC records, and the newer C2PA content credential markers, then hands back an image that looks the same but carries none of that history. Nothing is uploaded. The work happens on your own device, and the site makes that the whole point. The page's own summary is strip metadata off, keep image on, and that's the entire scope.

The problem is that photos say more than they show. A picture posted from a phone can carry the exact location it was taken, the device model, the editing software, and increasingly a C2PA manifest that records where the file came from and what touched it. C2PA in particular is spreading, since camera makers, editing tools, and platforms such as LinkedIn now attach content credentials to images, and most older stripping tools don't know the markers exist. Plenty of online strippers do exist, but most of them work by uploading your image to a server, which is an odd way to protect privacy. Meta-Stripper's answer is to never let the image leave the browser tab.

The method is re-encoding. Rather than trying to surgically cut metadata segments out of the file, it decodes the image to pixels and writes a fresh file from those pixels, which leaves nothing for the old blocks to live in. For JPEG that includes the APP11 segments where C2PA and JUMBF data sit, and for PNG the caBX chunks. You can drop in a batch of files at once, choose JPEG, PNG, or WebP as the output, and set the quality for lossy formats with a slider that defaults to 92 percent and is ignored for PNG. Because it re-encodes anyway, it can convert between the three formats on the way out, so a PNG screenshot can come back as a smaller WebP with its metadata gone in the same pass. After each image is written, a verification step re-reads the output and confirms that the known source metadata markers are absent, so you're not taking it on faith. The page states no batch or file size limit, so the practical ceiling is the browser's memory rather than a quota. Once you've visited the page it works offline as a progressive web app.

It fits anyone who shares images and would rather not share the circumstances of taking them. People selling items online from their own homes, privacy-minded users posting to social media, journalists handling photos from sources, developers who want a quick way to sanitise assets before they land in a public repository, and anyone who needs a clean copy of an image without its provenance chain for testing or comparison are the natural cases. It's a single-purpose utility, which is what most of those people want.

What makes it different from the crowd of metadata removers is the combination of C2PA awareness, local-only processing, and honesty about what it can't do. It states that it can't guarantee removal of perceptual fingerprints or invisible watermarks embedded in the pixels themselves, and that browser-generated technical encoding data may remain in the new file. That's a rarer admission than it should be, and it points to the C2PA specification's own soft binding mechanisms, which are designed to survive exactly this kind of re-encode. If you need to defeat those, this isn't the tool, and it says so. The privacy statements are unusually specific too, no server uploads, no accounts, no analytics, and no CDN, which means there's no third-party script loading alongside your photos.

The limits follow from its shape. Re-encoding means a lossy format is recompressed, so a JPEG goes through another generation at the chosen quality, and photographers who care about generation loss will want to keep the original. There's no command-line version, no API, and no information on the page about who maintains it, so you're trusting a small anonymous utility rather than a company. No licence is stated and no source repository is linked, so it can't be audited the way an open-source tool can, and if the page ever disappears there's no fallback beyond the installed copy. It handles three formats and nothing else, so HEIC, TIFF, and RAW files need converting first.

Access is free, with no account, no tracking, and no paid tier. Nothing on the page suggests a paid version is planned, and there's no donation link either, so the model appears to be a free utility offered as is. There's no contact address published, so support is whatever the page explains, which in fairness is a fair amount. If you'd rather not trust any web page with a sensitive image, the offline mode is the practical answer, since you can load it once, disconnect, and process the files with the network off. For a job this narrow, that's a reasonable bargain.