pictkura

pictkura does two things and stops: it imports from a USB drive or SD card into folders sorted by capture date, and it lays out a large library so you can find and look at what you already have.

It reads the NTFS change journal, so a second launch only visits files that changed since last time. Thumbnails are scaled with SIMD and JPEG is decoded at a reduced scale. Every search is an index seek (SQLite FTS5, with CJK expanded to bigrams). Tiles are sized from width and height held in the database, so nothing jumps around while a screenful loads.

RAW files are not developed — the display JPEG the camera embedded is used as-is, which keeps the camera's own colour and costs no demosaicing. 24 RAW extensions were confirmed against real files, including Apple ProRAW. HEIC, AVIF and video are handled too, and files that exist only in the cloud are never downloaded on their own: the real file is fetched only when you actually look at that photo.

What it deliberately does not do: no RAW development, no editing, no AI culling, no tagging, and one sort order only. Those omissions are listed openly on the site.

Free and MIT licensed, with no account and no telemetry. The only outbound connection is a once-a-day update check that can be turned off entirely.