Find exact duplicate photos or review visually similar images locally. Compare groups, exclude extra copies, and download a ZIP of the photos you keep.
🔒Processing stays on your device. Required libraries may download.
Zero Server Uploads
On-device Processing
Worked examples and how to read the result
Use exact matches for identical file contents and visual suggestions for manual review.
Keep one repeated download
Example input
Three photos: two byte-identical copies with different names, plus one distinct photo. Exact mode.
Expected result
One matching group. Excluding one copy produces a ZIP with two photos.
Different filenames do not affect content hashes. The original files remain on your device after export.
Visually similar results are approximate. Review the resolution and edits before excluding a suggested copy.
How Duplicate Photo Finder works
Find duplicate photos by selecting a local folder, including subfolders, or individual images and comparing SHA-256 hashes of the original file contents. Exact mode groups byte-identical files even when their names differ. Switch to visual similarity to look for re-encoded or resized copies, review the suggested groups, and choose which photos to exclude from a new ZIP.
Visual matching uses a sixty-four-bit difference hash calculated from a small grayscale representation, together with average color and aspect-ratio checks. It compares candidates with a reference photo for each group. The tolerance slider changes how many hash bits may differ. Similarity is a review aid, not proof that photos are interchangeable; edits, crops, rotation or plain backgrounds can produce missed matches or false suggestions.
Choose up to two thousand PNG, JPG or WebP photos, at most 50 MB each. Scanning processes one image at a time and stores small previews. ZIP exports are limited to 250 MB of kept originals. Every file is decoded locally, and unreadable files are listed as skipped. Checking Exclude from download affects only the new ZIP. The tool never deletes originals, changes a folder or frees disk space automatically. The kept-photo ZIP contains only successfully decoded files from this session.
How to use Duplicate Photo Finder
1. Choose a folder or photo batch
Select a folder and its subfolders or choose individual photos. Only PNG, JPG and WebP images count toward the two thousand photo limit; other files are skipped. Convert HEIC images to JPG first. The scanner reports progress while hashing and decoding the files.
2. Choose a matching method
Use Exact duplicates when you need byte-for-byte matching. Select Visually similar and adjust tolerance when looking for resized or re-encoded copies.
3. Review group members
Inspect filenames, thumbnails and file sizes. Select additional copies as a starting point, then adjust individual exclusions after reviewing the photos.
4. Download the kept set
Create a numbered ZIP containing unchecked photos and unmatched images. Skipped unreadable files are not included; keep them separately. No original files are deleted.
Key features and technical specifications
Content-based exact matching
SHA-256 compares file contents, not just names, dimensions or reported file sizes.
Adjustable visual suggestions
Combines image fingerprints, aspect ratios and average colors, with a tolerance control for review.
Non-destructive export
Exports original bytes of kept files with numbered names, keeping the originals and extensions intact.
Practical use cases
Review repeated downloads
Find identical photos saved under different names before collecting a cleaner export for a project.
Compare resized copies
Use visual mode to locate possible alternate-size versions, then decide which resolution and edit to keep.
Organize a shared batch
Download a reviewed ZIP for a colleague without granting the website access to delete or rewrite your photo folder.
Frequently asked questions
How do I find duplicate photos with different filenames?
Exact mode hashes each file’s contents. Two byte-identical files group together even if their filenames are different.
Does a visual match mean the photos are identical?
No. Visual matching is approximate and requires review. Different scenes can share similar fingerprints, while some edited copies can fail to match.
Will it delete duplicate pictures from my computer?
No. Exclusion only leaves a photo out of the downloaded ZIP. Originals are never deleted or overwritten.
Can I scan a folder and its subfolders?
Yes. Choose a photo folder to include PNG, JPG and WebP images recursively. Other file types are skipped automatically. Scan up to 2,000 supported photos, at most 50 MB each. Images are processed sequentially; ZIP downloads are limited to 250 MB of kept originals. Folder paths appear beside matches. Convert HEIC first; unreadable images are reported as skipped.
Which photos are included in the ZIP?
All successfully decoded, unchecked photos are kept, including unmatched images. Checked photos and skipped unreadable files are excluded.
Why do identical-looking photos not match in exact mode?
Different metadata, compression or edits change the file bytes and its hash. Try visual similarity, then inspect the suggested copies carefully.
Choose photos to compare
Drop files here or click to browse — processing stays on your device
Folder scanning includes subfolders and automatically skips other file types. Photos stay on your device. If your browser does not support folder selection, use the photo picker.
Scan up to 2,000 photos (50 MB each). Images are processed one at a time, with small previews to reduce memory use. Exact mode compares file contents. Similar mode compares a small visual fingerprint, color averages and aspect ratio; it can miss edits or suggest unrelated photos. Review every suggestion. Nothing is deleted from your device.