Unmarked
Remove the hidden data in your photos and PDFs, on your own device.
Unmarked shows you what a file gives away (where a photo was taken, which phone took it, who wrote a document and what
was deleted from it) and makes a clean copy without it. No account, no server, no upload.
Use it now: unmarked.anonacondacodelab.com · Android app:
unmarked.anonacondacodelab.com/android.html
Features
- Photos (JPEG, PNG, WebP): removes location, names, phone and camera details, serial numbers, dates and time zone,
embedded thumbnails (which can show what you cropped out), motion-photo video clips, edit history, document IDs,
comments and keywords. Lossless: the picture data is copied byte for byte; only the rotation and colour profile
are kept, so the photo still looks right. A standard sRGB profile is removed (colours stay the same); any other
profile keeps its colour data but loses its creation date, maker codes and copyright text.
- PDFs: removes document information (author, app, dates), XMP metadata, earlier saved versions that can hold
deleted text, page thumbnails, private app data, and camera data inside placed photos. Attached files, scripts and
comments (with their authors' names) are removed by choice.
- See before you clean: every finding is listed in plain language, most revealing first, marked Remove or Keep.
- Checked twice: each clean copy is inspected again before it's offered, and anything left is reported.
- Options: keep the date taken, keep copyright notices, give files plain names (
IMG_20261024_143512.jpg reveals the
date too).
- Many files at once, saved one by one or as a
.zip with neutral timestamps.
- Android app: Share → Unmarked from the gallery or any app; no internet permission and no other permissions.
How it works
The whole web app is one file, dist/index.html, built by node build.mjs from src/. A Content Security Policy
blocks every network request, so the page can't send your files anywhere.
The cleaning code is src/clean.js: self-contained, no DOM, usable in any browser or Node 18+ (other studio apps reuse
it). Images are parsed and cut down directly; PDFs use pdf-lib to rewrite the
document from its current version only.
Development
node build.mjs # builds dist/
npm test # builds, then runs the tests (Node's built-in runner, no dependencies)
Tests use real files in tests/fixtures/, made with ExifTool, Pillow, reportlab and pikepdf
(tests/fixtures/make-fixtures.sh rebuilds them). The Android app is in mobile/ (see mobile/README.md).
Hosting your own copy
Host the dist/ folder anywhere that serves https. The official site is on Cloudflare Pages, which publishes dist/
as committed (no build command), with the HTTP headers in dist/_headers.
Disclaimer
Unmarked is provided as is, without any warranty. It removes the hidden data it knows about, but no tool can
promise to find everything. Check clean copies before sharing anything sensitive. You are responsible for what you
share; Anonaconda Code Lab accepts no liability for any loss or harm from using this software.
Licence
Free software under the GNU General Public License v3 or later, with no warranty. © 2026 Anonaconda Code Lab.
Third-party components: see NOTICE.md. Security and privacy: see SECURITY.md.
Questions or problems: [email protected].
