Fast culling controls
Accept, reject, rate, tag, move, and delete from the keyboard without changing context.
Local-first desktop culling
Image Triage
Review huge folders, compare near-duplicates, mark keepers, and use optional local AI without handing your library to a cloud service.
Built for large folders
Virtualized thumbnails and background caching keep review sessions moving.
Decisions stay local
Ratings, tags, keep/reject state, and AI assets are stored on your machine.
Windows and Linux
MSI for Windows, AppImage for Linux, both published through GitHub Releases.
Review flow
Move from folder chaos to a clean keeper set.
Image Triage is tuned for the repetitive part of image work: opening a folder, scanning quickly, comparing similar shots, and preserving the choices you make along the way.
01Open
Load a working folder and keep navigation visible.
02Review
Scroll fast with cached thumbnails and async previews.
03Compare
Check exposure brackets, duplicates, and close alternates.
04Cull
Mark keepers, reject misses, rate, tag, move, or export.
Persistent folder memory
Warm-start folder scans and disk thumbnail caches help repeat sessions open faster.
Designed for inspection
Preview and compare modes keep EXIF details and file actions close to the image.
Optional local AI
AI help when you want it, normal review when you do not.
The app can install local runtime packages and model assets for DINO prefiltering, CLIP scoring, and TOPIQ quality checks. They are optional downloads so the core installer stays lean.
- DINO prefilterRanks and prepares candidates for AI-assisted review.
- CLIP ONNX bundleRuns local embeddings and category scoring.
- TOPIQ quality modelAdds technical quality signals during ingest.
Installers
Download the latest release from GitHub.
Use the Windows MSI for normal installs and updates. Use the Linux AppImage when you want a portable build that can be placed in your local applications folder.