Read locally, reported locally
This is the tool where the promise matters most, because the thing you are checking for is exactly the thing you would not want sent anywhere. The file is read with FileReader and parsed in JavaScript in the page. No request is made, no copy is retained, and there is nothing on a server to delete. Confirm it the direct way: open the network tab before you pick a file, or disconnect entirely and inspect a photo offline.
Format comes from the bytes
The extension is ignored and so is the MIME type the operating system reported. Every image format begins with a fixed signature — FF D8 for JPEG, .PNG for PNG, RIFF and WEBP for WebP — and that is what gets read. A PNG renamed to .jpg is identified correctly, which is worth knowing when an upload form rejects a file for reasons that make no sense.
Colour depth comes from the format's own header: PNG's IHDR chunk gives bit depth and colour type directly, and a JPEG's SOF marker gives sample precision and component count. Pixel dimensions come from the browser's decoder rather than from parsing, which is why a format the browser cannot draw shows a size but no dimensions.
What EXIF actually contains
EXIF is a TIFF structure tucked into a JPEG's APP1 segment, and cameras write a great deal into it. The capture timestamp to the second. Make and model of the body. Lens model on most interchangeable-lens cameras. Full exposure settings. If location services were on when the shutter fired, latitude and longitude to a precision that identifies a room, not a neighbourhood. Some bodies add the owner name field and the serial number, which correlates every photo from that camera even across accounts that share nothing else.
Only the fields people actually ask about are shown. A full dump runs to hundreds of tags including maker-specific blocks, and reading those properly is what ExifTool exists for.
Removing it, and where it gets removed for you
The reliable method is to redraw the image and save the result: a canvas holds pixels and nothing else, so encoding from one produces a file with no metadata at all. Every tool on this site that resizes, crops, rotates or compresses does exactly that, which means any of them strips EXIF as a side effect. The one exception is the Base64 encoder, which copies the original bytes and therefore carries the metadata straight through.
The large social platforms strip EXIF during upload, so a photo posted there is already clean. What does not strip it: email attachments, cloud storage share links, direct file transfer, most self-hosted forums, and anything that hands over the original. The safest habit is upstream of all of this — turn location tagging off in the camera app, and the coordinates are never written in the first place.
The orientation flag, explained once
Value 1 means the pixels are already the right way up. Value 6 means "rotate 90 clockwise when displaying", 8 means anticlockwise, 3 means 180. Phones almost always write 6 or 8 for vertical shots rather than rotating the pixels. Software that honours the flag shows the photo correctly; software that ignores it shows it sideways, and both are looking at the same file. If a photo is upright in some places and lying down in others, that is the flag, and re-saving it through the rotate and flip tool bakes the correct orientation into the pixels and drops the flag, ending the disagreement.
Questions people ask
It says GPS is attached but shows no coordinates.
A GPS directory exists in the file but the latitude and longitude tags inside it are absent, empty, or written in a form this parser does not accept. Treat that as attached until proven otherwise. Some cameras write a GPS block containing only a timestamp or a satellite fix status, but you cannot tell which from the outside, and the safe assumption before publishing is that something locational is in there.
Why is there nothing to show for my screenshot?
Screenshots have no capture metadata because no camera was involved — there is no exposure, no lens, no location. PNG can hold text chunks and some tools write a note into them, but this reads EXIF from JPEG only. A screenshot showing nothing is the expected result rather than a failure.
Can I edit the metadata, for example fix a wrong capture date?
This only reads. Writing EXIF means rebuilding the segment structure correctly, and getting it wrong produces a file that some decoders accept and others reject. ExifTool is the standard answer and handles the whole tag space including maker notes. It is a command-line tool, and worth the twenty minutes if you have a folder of photos with a clock that was set to the wrong timezone.
Does a file with no EXIF prove it was never geotagged?
It proves the EXIF is not there now. Anything that re-saved the image would have removed it, and that includes the platform it may have travelled through before reaching you. Absence tells you the current copy is clean, which is what matters for publishing it, but it says nothing about what the original held or where other copies of it went.