VisionTagger

Hi Product Hunt 👋 I built VisionTagger because I wanted my photo library to be truly searchable across thousands of images—with captions and keywords I can trust. I love how well Mac Photos handles face search, but Apple’s keyword-style search has been hit-or-miss for my library — and it’s frustrating that you can’t really see or manage the labels it relies on. I wanted a way to add reliable captions and tags to my photos without uploading anything to a cloud service. The idea started as “generate better captions” but quickly turned into “let me define the exact metadata I need”. So VisionTagger lets you use built-in sections (titles, descriptions, keywords, style, safety) and create your own custom fields (boolean/text/list) with per-field prompts — so the output matches your workflow and naming conventions. It runs fully on-device on Apple Silicon Macs on macOS 26 (Tahoe). After generation you can export XMP (via ExifTool) and JSON/TXT (per image or per batch), write back to your Photos Library, and/or apply Finder tags. There’s a free trial for 100 images. Curious: what metadata would make your library instantly easier to search or manage? And if you manage a large or professional library, what fields would you want to generate automatically?