Images · 100% Private Client-Side

Photo Duplicate Finder

Find exact and near-duplicate photos by hash, with a keep recommendation. Zero server uploads — all processing happens directly in your browser.

About this tool — how it works, limits & FAQ

What it does

Groups selected files into exact byte-duplicates (size + SHA-256, hashed in a Web Worker) and perceptual near-duplicates (aHash/dHash Hamming distance with a sensitivity slider and similarity percentage). Per-group thumbnail strips carry group id, size, date and hash columns with checkboxes to select copies, and the recommended keeper (oldest date, then shortest name) is pre-marked. Any file type is accepted.

Common use cases

  • Camera Roll Duplicate Pruning: Detect nearly identical photos taken seconds apart to reclaim valuable storage space.
  • Resized & Recompressed Photo Detection: Match identical images across folders regardless of differing dimensions or JPEG compression levels.
  • Burst Mode Photo Cleanup: Group rapid-fire photo bursts and identify the best candidate while flagging redundant extras.

How to use

  1. Select multiple files, or drop an entire folder — every file type is accepted, not just images.
  2. Exact duplicates are grouped by byte size then SHA-256 hashed; images also get aHash/dHash perceptual hashes.
  3. Review each thumbnail group, tick the copies to remove, or use Keep recommended to keep the oldest copy.
  4. Export the cleanup report as CSV (Group,File,Size,Date,Hash,Distance) or as a standalone HTML report.

Limits & considerations

Exact dedupe uses SHA-256 (needs https or localhost). Perceptual hashing catches resized and recompressed copies, but heavy edits or crops may fall outside the similarity threshold.

Frequently asked questions

How does perceptual hashing find duplicates at different sizes?

The engine uses dHash and aHash algorithms that analyze relative luminance gradients rather than exact file bytes, matching resized duplicates.

Does the tool automatically suggest which photo to keep?

Yes. The recommendation algorithm highlights the oldest original capture date or highest-resolution file as the keeper.

Are my private family photos uploaded to a cloud scanner?

No. All perceptual hashing and image comparison algorithms execute entirely inside your local browser memory.

Privacy & security

Hashing and thumbnailing run entirely in your browser with Web Crypto and Canvas. Your files are never uploaded to any server.

Related Images tools