Klyr for Mac · Exact duplicate detection

Find duplicate files by content, not by name.

Klyr narrows candidates by size, verifies their contents with SHA-256, and shows the paths for review. A duplicate result is evidence of an exact match—not permission to remove it.

Short answerKlyr finds exact duplicate files locally in two passes: same-size files become candidates, then SHA-256 hashes verify matching contents. Matching filenames alone are never treated as duplicates.

How duplicate verification works

Reading and hashing every file immediately would waste time on files that cannot possibly match. Klyr first groups regular files by their byte size. Only groups containing two or more same-size files advance to content hashing.

  1. Enumerate regular files in the folder you selected.
  2. Group candidates by size. Differently sized files cannot be exact copies.
  3. Calculate SHA-256 for candidates in matching-size groups.
  4. Report hash-equivalent groups and keep one member represented as the canonical copy.
Default minimumFiles below 1 KB are skipped because tiny duplicate configuration files usually offer negligible recovery.
Local processingFile contents and hashes are computed on the Mac and are not uploaded for duplicate scanning.
CancellationThe scan checks for cancellation during enumeration and hashing.
Exact duplicates onlySimilar photos, edited media, and same-name documents with different contents are not labeled exact copies.

Why every duplicate still needs review

Two files can be byte-for-byte identical and both still matter. One may belong to a backup, a synchronized folder, an application bundle, or a project workflow that expects a particular path. Klyr shows duplicates for human review instead of automatically choosing a deletion policy.

The normal removal path sends only selected copies to macOS Trash after path, whitelist, and sensitive-location checks.

APFS clones and reclaimable space

APFS can create logical copies that initially share physical blocks. Two files may have identical contents and each report a full logical or allocated size while still sharing storage underneath. Removing one clone may therefore recover less physical capacity than its displayed size suggests.

Klyr treats duplicate identity and real-world reclaimable capacity as related but different questions. Rescan the volume after removal to confirm the result.

Duplicate finder questions

Are same-name files necessarily duplicates?

No. Names are labels, not content proof. Klyr verifies same-size candidates using SHA-256.

Can Klyr find similar photos?

Klyr’s exact duplicate scanner detects byte-identical files. A resized, recompressed, cropped, or edited photo is not the same exact file and should not be represented as one.

Will every reported duplicate free its full displayed size?

Not always. APFS clones and shared blocks can reduce the physical space recovered after one copy is removed.