My iPhone library grew to thousands of photos, and many of them were basically the same shot: multiple takes of the same pose, bursts, and old screenshots.
Most duplicate cleaners I tried only caught exact pixel duplicates, so I still had to scroll through everything myself.
SnapSift scans the library on-device, groups photos that look alike, and helps you keep one best shot from each group.
- Scans the whole library in under a minute
- Shows similar photos side by side
- Picks a likely “best” photo in each group
- Lets you delete the extras quickly
Everything runs locally using Apple’s Vision framework. Photos never leave the device, there is no server and no cloud.
You can scan for free and delete up to 25 photos per month on the free tier. Pro unlocks unlimited cleanup.
I would like to hear how well the grouping works on your library, and any cases where the “best pick” is clearly wrong.
I built this because I kept exceeding my mobile data cap when tethering.
The problem: macOS doesn't distinguish between home WiFi and iPhone hotspot. I'd finish a Zoom call thinking I was on home internet, only to discover I'd burned through 3GB of cellular data.
Existing options didn't fit:
- TripMode ($15/year) does app blocking - overkill for just monitoring
- Bandwidth+ is free but treats all WiFi the same
Technical approach:
- NWPathMonitor.isExpensive detects tethering vs regular WiFi
- CoreWLAN reads SSID for multi-hotspot profile support (requires Location permission on Sonoma+)
- Pure system frameworks, no external dependencies
Supports multiple profiles so you can track personal phone (15GB cap), work phone (unlimited), travel MiFi separately.
$4.99 one-time. macOS 14+.
Happy to answer questions about the implementation.