What does the EXIF Remover do?
It reads the block of hidden tags your camera wrote into a photo file, and hands back the same photo with those tags gone. The tags are the interesting part. A typical phone picture carries the GPS coordinates of the spot you stood on, the second you pressed the shutter, the make and model of the phone, often the camera body's serial number, the lens, and the name of the last app that touched the file. None of it shows up in the picture. All of it travels with the file.
Drop a JPG, PNG, or WebP on the page and two things appear at once — a list of what was actually hiding in that file, and a clean copy already built. There's no Strip button, because pressing one was never the reason you came. Read the list, click Download clean copy, done. Up to 50 MB per file, up to 20 files at a time, and a batch comes down as a ZIP.
One detail matters more than the rest: the photo never leaves the tab. Parsing and rewriting both happen in your browser, on bytes your device already has. There's no upload, no queue, no account, and nothing sitting on someone else's disk afterwards.
What a photo says about you that you didn't type
EXIF — Exchangeable Image File Format — is a structured block cameras write alongside the image data. Two neighbours travel with it: XMP, which can hold edit history and contact details, and IPTC, which holds captions, credits and copyright. Here's what's usually in there.
| Tag | What it gives away |
|---|---|
| GPS latitude / longitude | Where you were standing, usually accurate to a few metres. On a photo shot at home, that's your address. |
| Date and time taken | The exact second, often with a timezone offset — enough to place you somewhere at a moment. |
| Camera make and model | Which phone or body shot it. |
| Body serial number | The quietly interesting one. Two photos posted from two unconnected accounts can be tied to the same physical camera by that number alone. |
| Lens | Focal length and aperture of the specific lens used. |
| Software | OS version, and whichever editor last saved the file. |
| Owner and copyright | A real name, usually typed into a camera's settings once, years ago, and forgotten. |
Most people meet this the same way: they're about to post a picture to a classifieds listing, a forum, a dating profile or a support ticket, and it occurs to them that the file might know where their house is. It usually does. Removing the GPS from a photo before it goes out takes about ten seconds, and the EXIF Remover shows you the coordinates on the way past so you know what you just deleted.
How the EXIF Remover works
Lossless by default
Nearly every tool that strips EXIF data from a photo does it by decoding the image and saving it again. The tags disappear — as a side effect of throwing your pixels away and recompressing them. Do that two or three times and the photo visibly degrades. You wanted this photo minus the tags, not a slightly worse copy of it.
So this one doesn't re-encode. It walks the file's container structure, drops the segments that carry metadata, and copies the compressed image data through byte for byte. A JPEG that needs no rotation comes back pixel-identical to the one that went in — not "visually identical", identical. In the example further down, all 4,400,011 bytes from the JPEG's scan marker to the end of the file match the original exactly.
What each format keeps, and what goes
| Format | Removed | Kept |
|---|---|---|
| JPEG | EXIF and XMP (APP1), IPTC and editing blocks (APP13), comment segments | The JFIF header, the embedded ICC colour profile, and every byte of image data from the scan marker onward |
| PNG | eXIf, the tEXt / zTXt / iTXt text chunks, tIME, and every other non-essential chunk | Image, palette, transparency, gamma and colour chunks, the ICC profile, and the animation chunks that keep an APNG animating |
| WebP | The EXIF and XMP chunks — plus the container's flag bits get cleared and its size field rewritten | The image bitstream, alpha, ICC profile, and animation chunks |
PNG uses a whitelist rather than a blocklist, which is the stricter choice: an unknown chunk can carry anything, so anything not on the keep list goes. The colour profile survives on purpose — a profile is rendering data, not personal data, and dropping a Display P3 profile makes a photo look washed out on every screen that honours it.
It checks its own work
After building the clean copy, the tool re-scans those exact output bytes with a second, separate pass — including a raw sweep of the region after a JPEG's scan marker, where an appended block can hide from a normal walk. Only if that pass finds nothing does the page say No metadata left. If something survived, it says so by name and still offers the file for download, because hiding a file it couldn't fully clean would be the worse answer. A green line that can't ever go red isn't evidence.
A worked example, start to finish
Take a JPEG called IMG_2041.jpg carrying a full phone EXIF block. Drop it on the page and this is the readout, word for word:
- GPS: 37.8024, -122.4058 — with a Copy coordinates button, so you can paste it into a map and see for yourself what the file was handing out
- Taken: 2026:04:11 18:22:07
- Camera: Apple iPhone 14 Pro (serial ending 7K2)
- Lens: iPhone 14 Pro back camera 6.86mm f/1.78
- Software: 17.1.1
- Owner: Dana Reyes · Copyright: © Dana Reyes
Summed up in one line: "Found GPS location, Taken, Camera, Lens, Software, Owner, Copyright and 1 other tag." Underneath, in green: "No metadata left." The download button saves IMG_2041-clean.jpg — same format, same resolution, same pixels, minus every tag above.
A small WebP shows the size arithmetic more clearly, because the numbers are small enough to see: a 58-byte file carrying a 16-byte EXIF chunk comes out as listing-clean.webp at 42 bytes, labelled "58 B → 42 B · 16 B of metadata removed". On a real photo the proportions flip — the metadata is a rounding error against the image data, which is exactly why nobody notices it's there.
Want to check the no-upload claim rather than take it on faith? Open your browser's network tab, then drop a photo. Nothing goes out. That's a thirty-second test anyone can run, and it's the right instinct to have about a page that handles a file you're trying to keep private.
When a photo actually leaks
Most large social networks strip metadata when you upload — which is exactly why the cases that matter aren't social networks. The file passes through untouched when you attach it to an email, send it in a direct message, upload it to a marketplace or classifieds listing, post it on a forum, or drop it into a small site's contact form or a support ticket. Those are the routes where a photo taken at the kitchen table quietly carries the kitchen table's coordinates.
The listing photo is the classic. You photograph a bike in your driveway, post it to sell, and the ad now contains the driveway's latitude and longitude for anyone who cares to look. Stripping first costs you about ten seconds and removes the question entirely.
Which brings up the thing that made this tool worth building. Search for a way to remove metadata from a photo and nearly every result asks for the same thing: upload the exact file you're trying to keep private to a stranger's server, so that server can delete a GPS tag on your behalf and hand you back a link. That's a strange trade, and it's the default across the category. The alternatives aren't better: a Windows-only executable you have to install, or metadata removal buried three menus deep inside a photo editor that wants an account, a subscription, and a fourteen-day trial before it will delete a tag for you. None of that is a technical necessity. Reading and rewriting a photo's container is a few hundred lines of byte arithmetic that a browser can run in a few milliseconds. The paywall was a business model, not a limitation.
So this runs on your machine, in the tab, for anyone — the freelancer listing gear, the student posting to a class forum, the person selling a sofa, the Fortune 500 employee cleaning up a photo on a Saturday. Same tool, same speed, no tier.
Where this goes wrong, and what to check
The sideways-photo trap
Phones often store a photo in the sensor's raw sideways frame plus an EXIF Orientation tag telling the viewer to turn it. Delete the tag without turning the pixels and the portrait comes back rotated 90 degrees — a very common failure elsewhere, and an annoying one, because the result looks fine in the tool and wrong everywhere else. When Orientation is anything other than upright, the EXIF Remover bakes the rotation into the pixels and says so plainly: "This photo was stored sideways with a rotation tag. We turned the pixels the right way up, so this one was re-saved." That single file is re-encoded; every other photo in the batch takes the lossless path.
HEIC, PDFs, and formats that don't fit
iPhones shoot HEIC by default, and browsers can't open it. Drop one and you'll get a pointer to the HEIC to JPG converter rather than a silent failure — convert, then come back. Drop a PDF and you'll be pointed at the PDF metadata editor, which edits document properties instead of stripping them. GIF and TIFF aren't supported yet. Format is read from the file's first bytes, never its extension, so a mislabelled .jpg gets handled correctly either way.
What stripping doesn't do
Removing the tags removes the tags. That's real and useful, and it's also the whole claim. It doesn't touch what's inside the picture — a street sign, a house number, a reflection in a window, a school uniform. It doesn't address sensor-level fingerprinting, where per-camera noise patterns can in principle link images. And a corrupt or truncated file gets refused outright rather than half-cleaned; you'll see "We couldn't read this image. It looks damaged." and no download, because a partial file built from a bad offset is worse than none. Treat the EXIF Remover as closing the obvious door, not as anonymity.
Related image tools
Stripping metadata is usually one step in getting a photo ready to post. The image compressor brings the file size down for a listing or an email attachment. The image cropper and image resizer handle framing and dimensions. If the picture itself shows something it shouldn't — a face, a plate, a document on the desk — the blur tool deals with the pixels the way the EXIF Remover deals with the tags. And JFIF to JPG fixes the odd file extension that some browsers hand you on download.
Every one of those runs in your browser, with no account and no upload, and 10% of every dollar Microapp earns goes to charity — off the top, audited quarterly.
Frequently asked questions
Does the EXIF Remover change the photo's file date?
Removing the EXIF data deletes the capture date stored inside the file, yes — that's one of the tags. But the download is a brand-new file, so your operating system stamps it with the moment you saved it. If you're organising a library and you need the original capture date, copy it out of the readout before you download; the tool shows it to you on screen precisely so it isn't lost silently.
Does my photo's colour look different afterwards?
No. The embedded ICC colour profile is deliberately kept, on JPEG, PNG and WebP alike. Profiles get dropped by a lot of strippers, and the symptom is a photo that suddenly looks flat or washed out on a wide-gamut screen. A colour profile describes how to render the picture; it says nothing about you.
Will an animated PNG or WebP still animate?
Yes. The animation chunks are on the keep list for both formats, and for WebP the container's size field is rewritten and its flag bits cleared so strict decoders don't go hunting for chunks that are no longer there. Getting that size field wrong is the silent failure mode for WebP — the file looks fine until something rejects it.
How many photos can I clean at once?
Twenty files per batch, 50 MB each, 200 MB total. Drop more than twenty and the first twenty are taken with a note rather than an error. When there's more than one clean copy ready, Download all packages them into a single ZIP. Files are read one at a time rather than all at once, which is what keeps a large batch from exhausting memory on a phone.
Does this work on a phone?
Yes — it's a web page, so a mobile browser runs it the same as a desktop one, and the photos stay on the handset. Pasting from the clipboard works too, which is the quickest route for a screenshot you're about to send.
How much smaller will the file get?
Usually not much, and that's the point. Metadata is typically a fraction of a percent of a photo's bytes — a few hundred bytes to a few dozen kilobytes against several megabytes of image data. The page shows the exact before-and-after so you can see it. If you actually need a smaller file, strip the metadata here and then run it through the image compressor; the two jobs are separate, and mixing them is how you end up with a re-compressed photo you didn't ask for.