I’d like to recommend a local photo management tool: Lap. It’s designed for family photo albums, photography enthusiasts, and large local media libraries starting at 100,000 photos. It doesn’t use the cloud, doesn’t force uploads, and doesn’t lock you into its system—all browsing, searching, organizing, and deduplication are done right on your own computer. Shortly after its launch, it garnered over 3.7k stars on GitHub, making it one of the hottest new projects in local photo management recently.
What problem does it solve?
Cell phones, cameras, and trips—you end up with a pile of duplicate photos, and when you want to organize them, you can’t find a tool that works for you. This is a real dilemma for many people. Traditional cloud photo albums either require you to upload private photos or lock your files in their own proprietary formats. Lap’s approach is straightforward:Your photos are simply regular files on your hard drive; the software only creates an index and organizes the data externally—it never moves the original images into a closed database.. You can still manage the same set of folders as usual in File Explorer or Finder.
Key Features
1. Truly offline, with privacy as a top priority. All AI processing runs locally, with no account required and no mandatory cloud synchronization, thereby avoiding the risk of privacy leaks associated with cloud photo albums.
2. Folder priority, no lock-in. Simply point the software to an existing photo directory to create a database; the software generates thumbnails and a search index without modifying the original files. It supports multiple databases, drag-and-drop import, file system synchronization, and secure moving, copying, and deleting.
3. Local AI Search. Supports natural language text-to-image search, visual similarity search, subject recognition, and face clustering, with optional multilingual search (covering 50+ languages)—all available offline. On the left side, you can also quickly browse by calendar, people, locations, cameras, and shots.
4. Intelligent Categorization and Duplicate Removal. The database automatically organizes photos by themes such as children, pets, portraits, food, sports, and landscapes; right-click on duplicate or similar photos to find related photos or similar people, clean them up in bulk, and see how much storage space will be freed up.
5. Da Ku is silky smooth. The official goal is to enable smooth browsing and zooming of over 100,000 photos. In actual testing, photos in formats such as JPG, PNG, and RAW (e.g., Canon CR3) can be previewed quickly, with no lag when zooming in or out.
6. Wide range of supported formats. Supports over 60 photo, RAW, and video formats; recognizes Apple Live Photos and Android Motion Photos, and handles RAW+JPEG/HEIC pairs and Live Photo components together when renaming, moving, or deleting them.
7. Practical Organization and Editing. Smart Albums (rule-based views), Collections (group photos without moving the original files), a four-pane comparison selector, built-in cropping, rotating, flipping, zooming, and basic color correction, as well as a map view for browsing photos with GPS data.
8. Multi-platform. Compatible with Windows, macOS, and Linux; the interface supports nine languages, including Simplified Chinese.
Target Audience
Privacy-conscious users who don’t want to entrust their family photo albums to third parties; those who manage the same set of folders in Finder or File Explorer and want quick access to their files; and photographers who need to quickly select and deduplicate photos across large RAW libraries.
Why It's Worth Paying Attention To
Photo management has long been a common need, but the "local-first + local AI" approach didn’t truly take off until 2026—as on-device models became more powerful and Rust/Tauri made desktop apps both lightweight and fast, Lap happened to be right on the cusp of this trend. Plus, it’s fully open-source under the GPL-3.0 license, offers interfaces in nine languages, and runs across three platforms, making it user-friendly for both beginners and developers. The project is still young (about two years old, with 1.6k more stars in the last 30 days), actively evolving, and backed by clear documentation.