Solo shots for covers
Filter '1 person' and type a concept like 'beach at sunset'. Close-ups with a single subject surface immediately, which is exactly what a YouTube cover or a reel thumbnail needs. Details on the face recognition page.
You publish every week, and your Mac holds thousands of shots, exports and screenshots from past shoots and campaigns. Today the brief asks for a vertical solo shot for a YouTube cover, and it is buried somewhere under files named final_v3_export.png. PhotoCore reads the content of every image on device: filter '1 person', type 'beach at sunset', and the frame is on screen, with search answering in under 50ms across 10,000 images.
A content library grows in every direction: camera shoots, phone captures, brand assets, screenshots of analytics and slides. The shot you need exists, you shot it in March, and its filename says nothing. PhotoCore indexes those folders in place and lets you ask for the image the way you remember it.
Filter '1 person' and type a concept like 'beach at sunset'. Close-ups with a single subject surface immediately, which is exactly what a YouTube cover or a reel thumbnail needs. Details on the face recognition page.
Switch the same filter to 'more than 3 people' when the post is about the team, a workshop or a collaboration. Combine it with text, like 'more than 3 people' plus 'around a whiteboard', and skip the solo portraits entirely.
One toggle hides screenshots and documents when you need real photos. Flip it back when you need that one slide: screenshots are indexed by content too, so 'pricing slide with a blue chart' finds it. See photo culling.
Sort any result by on-device quality scoring before you export. The sharp frames rise to the top, the soft ones sink, and you pick the winner yourself in seconds instead of pixel-peeping ten near-identical shots.
AI rename looks at each image and proposes a filename that describes it, using a local model by default. A folder of anonymous exports becomes a folder you can read at a glance, and the files stay findable in Finder, in your uploader and in every tool downstream. If you prefer, point it at a remote endpoint with your own key, stored in the macOS Keychain.
Search runs on SigLIP 2 through Core ML on the Apple Neural Engine: under 50ms across 10,000 images, in more than 100 languages, across JPEG, PNG, HEIC, TIFF, WebP, GIF and BMP. The full detail is on the AI photo search page. Everything runs on device, works offline, and keeps working when an external drive is unplugged, so the content library travels with the Mac.
Download PhotoCore free and index up to 200 images to try search, people-count filters, the screenshots toggle and quality sort. Full access is a one-time purchase of 59.99 euro, or 29.99 euro Early Bird until July 15.