Facet
A self-hosted AI photo manager that scores and culls shots, recognizes faces and searches photos in plain language, with CPU or optional GPU processing.
Facet combines photo culling with a self-hosted gallery for photographers choosing keepers from a shoot or organizing a personal library. Its AI analysis runs on your own machine without cloud services, accounts or API keys.
- Quality ranking: Compare photos by composition, exposure, sharpness and face quality, with scoring weights based on the photo's category; A/B comparisons train the personal "My Taste" ranking.
- Culling: Group bursts and similar shots, inspect per-face blink and expression badges, and preview automatic keeper decisions while preserving exposure brackets and panorama or HDR sets.
- Search and organization: Find photos with plain-language queries, group and name people with face recognition, and browse by timeline, map, folder or album, including smart albums built from saved filters.
- Editing workflow: Exchange standard sidecar files with Lightroom, darktable and digiKam, and preview RAW photos through your own darktable styles when judging which shots to keep.
- Uploads and sharing: Receive phone photos through the WebDAV inbox using PhotoSync or another compatible app, and create client proofing links or a portfolio site from an album.
Facet runs through Docker or natively. A laptop CPU is enough; an optional GPU accelerates analysis. It accepts JPEG, HEIC, camera RAW files and other still-image formats. When a RAW file can't be decoded, it can score a sufficiently large embedded preview instead. The local culling workflow is free.
