Open a local photo folder and clean it up by duplicates, similar shots, blurry frames and file size — free, with an automatic backup before anything is deleted.
Try it Free →The folder is read straight from your disk with the File System Access API. Thumbnails, hashes and EXIF are all computed in your browser — nothing is ever uploaded.
Files are bucketed by size and hashed with SHA-256, so identical copies are grouped with certainty. The earliest shot in each group is kept and the rest are pre-selected for you.
An 8x8 dHash plus Hamming-distance clustering groups burst shots and lightly edited versions, and marks the sharpest, highest-resolution frame as the one to keep.
A Laplacian-variance sharpness score ranks photos from blurriest to sharpest, with extra flags for tiny resolutions and abnormally small files.
Every file is copied into a timestamped _pudone_backup folder before it is removed. One click restores an entire batch to its original paths.
Hashing and analysis run in Web Workers with a concurrency pool, thumbnails are virtual-scrolled, and results are cached in IndexedDB so a second scan only touches new or changed files.
Click "Choose photo folder" and grant access. The scan starts immediately and photos appear as they are processed — you do not have to wait for it to finish.
Switch between Timeline, Duplicates, Similar, Blurry, Large files and All photos. Each view supports sorting, filtering by name, size, date and file type, and keeps your selection across tabs.
Click to select, Shift-click for a range, or drag a box across the grid. In duplicate and similar groups, "Keep this, drop the rest" selects everything else in one click.
Confirm the count and total size. Files are copied into _pudone_backup first, then removed. Use "Undo & restore" or the Backups drawer to bring any batch back.