Duplicate Image Finder
Find identical or visually similar images among a batch you upload.
About this tool
Select several images and this tool finds ones that are identical or visually very similar — handy for cleaning up a photo folder full of near-duplicates (screenshots taken twice, burst-mode shots, re-saved copies at different sizes). Comparison uses a perceptual "average hash" of each image's overall look, not a byte-for-byte file comparison — so a resized or slightly re-compressed copy of the same photo is still correctly matched.
Everything runs locally in your browser using the canvas element — no images are uploaded anywhere.
Frequently asked questions
- What does "similarity sensitivity" do?
- At "Exact matches only", only images with an identical average-hash fingerprint are grouped together. Raising it groups images that merely look alike (similar composition, colors, and brightness), even if they aren't literally the same photo — useful for catching near-duplicates, but can also produce false positives at high settings.
- Will this catch a photo that's been cropped or rotated?
- Not reliably — the average-hash method compares overall brightness patterns on a fixed grid, so cropping, rotation, or flipping changes that pattern enough that even very similar photos may not match. It's best at catching resized, re-compressed, or lightly edited copies of the exact same shot.
- Is any image uploaded anywhere?
- No — hashing and comparison both happen entirely in your browser.