Guides · Updated 2026-09-24

How to find duplicate photos on Flickr

Flickr does not detect duplicates, on upload or afterwards, and its website has no view that lists them. The same file uploaded twice sits in your photostream as two photos with two sets of views and favourites. Finding them by eye in a library of thousands is not realistic; finding them by the data Flickr does expose is, and takes one scan.

Step by step

  1. 1. Know what Flickr will and will not tell you

    Flickr's API gives no file size and no checksum for a photo, so a duplicate cannot be found the way a disk tool finds one. It does give each photo's original dimensions and the exact second it was taken, from the camera's own data. Two photos that agree on both are the same exposure in any real library.

  2. 2. Scan the library once

    In Photo Stream Assistant, open Library health and start the scan. It reads the whole library in pages of 500, which is a few minutes on a large account, and reports a count of possible duplicates alongside the other checks. The report is kept on your Mac, so you can come back to it without scanning again.

  3. 3. Open the count and look at the pairs

    Every member of a group is listed, not only the extra copies, so you see both photos side by side: their titles, whether one is in an album or has comments and favourites. Nothing has been decided for you.

  4. 4. Delete the copy you do not want

    Select the copies to remove and press Delete; the confirmation names the count. Deleting a photo on Flickr also removes its comments and favourites, so keep the one that has them. The deleted photos leave the list as you go.

  5. 5. Stop the next one at upload

    The app's upload checks each file's title against your library before sending it and flags a whole-title match as already on Flickr, with a tick to send it anyway. The website's uploader does not check.

Questions

Can Flickr find duplicates itself?

No. Neither the website nor the API has a duplicate check, on a library or at upload time.

Why dimensions and the date, and not the image itself?

Because those are what the API exposes for every photo in one read. Comparing the pixels would mean downloading every photo, which for a large library is hours and is throttled by Flickr. Dimensions to the pixel plus the exact second is a strict enough test that false pairs are rare, and every pair is shown for you to judge.

What about photos with no date?

They are left out of the duplicate check rather than lumped together. A scan or a screenshot has no capture time, and ten of them at the same size are not ten copies of one photo.

Is this part of the free app?

Library health, including the duplicate check, is part of the one-time Pro Unlock, because it reads and groups the whole library. Deleting is free and goes through the same confirmation as anywhere else.

In the app

More guides

Written by Stéphane Monier, who built Photo Stream Assistant for a library of 138,000 photos. Also available as plain Markdown.