About PrivacyMetaData

1. Our mission

Every photo your phone takes quietly records more than the picture: the camera model, the exact minute of the shot, often the serial number of the device, and — if location services were on — GPS coordinates accurate to a few metres. Most people have never seen this data, never agreed to share it, and have no easy way to remove it.

Our mission is narrow on purpose: make removing that hidden data so simple and so trustworthy that there is no reason to skip it. No account, no payment, no software install, and no upload. If a privacy tool asks you to hand over the very file you are trying to protect, something has gone wrong in its design.

2. Why we built it this way

The defining decision was to do all the work on your device rather than ours. Modern browsers can read a file from disk with the File API, parse its bytes in memory, and write a brand-new image to a canvas element — no server required. That one choice removes an entire category of risk:

• Nothing to intercept. Your image bytes never travel across the network to us, so there is no upload to be captured in transit.
• Nothing to breach. We hold no library of your photos, because we never receive them. A breach of our hosting would expose no user images.
• Nothing to trust us about. You do not have to believe a promise that we delete files. We cannot delete what we never had.
• Nothing to wait for. Cleaning happens at local speed, with no queue and no upload time.

Closing or refreshing the tab discards everything that was in memory. This is also why there is no sign-up: an account would create personal data where none needs to exist.

3. How the tool actually works

We think you should know the mechanism rather than take "metadata removed" on faith. There are two separate steps.

Reading your file

Our parser inspects the raw bytes of your image and reports every metadata block it recognises. For JPEG that includes EXIF (camera, lens, timestamps, exposure and the GPS sub-directory), XMP, IPTC, ICC colour profiles and plain JPEG comments. For PNG it reads the text chunks and embedded EXIF; for WebP, the EXIF, XMP and ICC chunks. GPS values stored as degrees-minutes-seconds are converted to decimal so you can see exactly where the photo places you.

Writing the clean copy

The cleaned file is not an edited version of your original with the fields blanked out. Your browser decodes the image to raw pixels, draws those pixels onto a canvas, and encodes a completely new file from them. Metadata containers are therefore absent rather than emptied — there is no leftover EXIF block to recover, because none was ever written.

That approach has one honest trade-off: JPG and WebP are re-compressed when they are rebuilt, so there is a very small quality change. We save at a high quality setting, and PNG output stays lossless. Our disclaimer sets out the limits in full.

4. We tested our own claims

Rather than assert that re-encoding works, we measured it. We built a single JPEG carrying a known payload of 18 readable EXIF fields — camera make and model, lens, body serial number, two timestamps, software, artist and full GPS coordinates — then pushed it through five different pipelines and re-parsed each output.

Every pipeline removed all 18 identifying fields. Two outputs gained one field that was not in the original: an ICC colour profile added by the browser while encoding. It describes colour, not you, and we report it because a tool that claims "zero metadata" without qualification is overstating its case. The full method and results are in our guide to metadata privacy and how to protect yourself.

5. Who this is for

The tool is deliberately general, but a few groups rely on it most:

• Anyone selling online. Marketplace photos taken at home can publish a home address that was never typed into the listing.
• Photographers. Sharing a portfolio image without exposing a camera serial number or the exact location of a private shoot.
• Journalists and researchers. Protecting where a photograph was taken, and by extension who was there.
• Job seekers. Documents and images attached to an application often carry an author name and editing history.
• Everyday social sharing. Platforms strip some fields from the public copy, but the original upload still reaches their servers intact.

6. What we do not do

Being clear about the boundaries matters as much as listing features.

• We do not use artificial intelligence. The cleaning is a deterministic decode-and-re-encode, and we describe it that way rather than dressing it up.
• We do not store, queue, log or analyse your images, because they never reach us.
• We do not ask for your name, email address or payment details.
• We do not claim the tool anonymises the content of a picture. Street signs, house numbers, reflections and faces can still identify a place or person.
• We do not promise that features labelled "coming soon" will arrive by a particular date.

7. How the site is funded

Advertising pays for the hosting, and that is the whole business model. There is no subscription, no premium tier, no watermark on your output and no cap on how many files you clean. We do not sell data — we have none to sell, since the tool collects no files and the site has no accounts.

Our privacy policy explains precisely what the website itself records, including server logs, aggregate analytics and the advertising cookies our partners may set, along with the relevant opt-out links.

8. Who writes this site

PrivacyMetaData is built and maintained by Suzon Mahmud. The metadata parser, the cleaning pipeline and every article here are his work, including the five-pipeline survival test described above. Where a claim comes from an outside authority — a regulator, a standards body or a privacy organisation — we name and link the source rather than asserting it ourselves.

If something on this site is wrong, unclear or out of date, we would genuinely like to know. Corrections are welcome and we update pages rather than leave stale advice in place.

9. Contact us

Questions about the tool, a bug report, a correction, or a press or partnership enquiry — email [email protected] and we will reply.

10. Related pages

To go deeper: our privacy policy covers what the site collects, our terms of use set out acceptable use, and our disclaimer explains the tool's format limits. The blog holds our longer guides, starting with metadata privacy.