FaceTagger — A Modern Local Alternative to Windows Live Photo Gallery & Picasa
If you've been searching for something to fill the gap left by Windows Live Photo Gallery (discontinued 2017) and Google Picasa (killed 2016), FaceTagger was built with exactly that loss in mind. Like those tools, it handles large local photo libraries with face detection and tagging — but it does it with AI that would have seemed like science fiction back then. There are no subscriptions, no cloud uploads, no account required, and no photo data ever leaves your machine. All AI models, databases, and settings live inside the application folder, so it's fully portable: put it on a USB drive or a network share and it works exactly the same.
What makes FaceTagger stand out against other modern alternatives is the depth underneath what looks like a straightforward face tagger. The AI suggestion engine continuously scores untagged faces against your confirmed identities in real time, surfacing its best guesses as you work rather than making you do manual batch runs. When you confirm a face, that decision gets written directly into the image file's metadata (or a sidecar file for RAW formats like NEF and DNG) using ExifTool — so your tagging work survives a complete database wipe and is readable by other software. Quality scoring automatically filters out low-value detections like ears and blurry blobs before they clutter your view, something neither WLPG nor Picasa ever managed. And because it's built to handle truly large libraries without freezing, it uses memory-mapped processing, WAL-mode SQLite, and a GPU-aware inference pipeline that keeps the UI responsive while the AI scans in the background.
Linux users in particular have been badly underserved here. Most face-tagging tools either don't run on Linux at all, or treat it as a secondary platform — a port that was done reluctantly, tested minimally, and left to bitrot. FaceTagger is one of the first face taggers built for Linux as a genuine first-class target, not an afterthought. Both the Linux and Windows builds go through the same development and testing, both get the same features at the same time, and the Linux version is not a slower or stripped-down cousin of anything. If you migrated from Windows to Linux and assumed you'd have to give up proper photo face-tagging, FaceTagger is the answer you've been waiting for. Whether you're migrating a decade of family photos or starting fresh, it's the closest thing to "what Photo Gallery should have grown into" — and unlike Photo Gallery, it actually runs on the OS you're using now.
DOWNLOADS:
Windows: (ppApp only for now):
Linux: (LLApp only for now):
Please if you use this and it works or doesn't let me know so I can fix any issues, just reply here and say how it goes. I'll do a full on release soon, but I am pretty worn out polishing the last few bugs getting windows to work as well as Linux was.
If you've been searching for something to fill the gap left by Windows Live Photo Gallery (discontinued 2017) and Google Picasa (killed 2016), FaceTagger was built with exactly that loss in mind. Like those tools, it handles large local photo libraries with face detection and tagging — but it does it with AI that would have seemed like science fiction back then. There are no subscriptions, no cloud uploads, no account required, and no photo data ever leaves your machine. All AI models, databases, and settings live inside the application folder, so it's fully portable: put it on a USB drive or a network share and it works exactly the same.
What makes FaceTagger stand out against other modern alternatives is the depth underneath what looks like a straightforward face tagger. The AI suggestion engine continuously scores untagged faces against your confirmed identities in real time, surfacing its best guesses as you work rather than making you do manual batch runs. When you confirm a face, that decision gets written directly into the image file's metadata (or a sidecar file for RAW formats like NEF and DNG) using ExifTool — so your tagging work survives a complete database wipe and is readable by other software. Quality scoring automatically filters out low-value detections like ears and blurry blobs before they clutter your view, something neither WLPG nor Picasa ever managed. And because it's built to handle truly large libraries without freezing, it uses memory-mapped processing, WAL-mode SQLite, and a GPU-aware inference pipeline that keeps the UI responsive while the AI scans in the background.
Linux users in particular have been badly underserved here. Most face-tagging tools either don't run on Linux at all, or treat it as a secondary platform — a port that was done reluctantly, tested minimally, and left to bitrot. FaceTagger is one of the first face taggers built for Linux as a genuine first-class target, not an afterthought. Both the Linux and Windows builds go through the same development and testing, both get the same features at the same time, and the Linux version is not a slower or stripped-down cousin of anything. If you migrated from Windows to Linux and assumed you'd have to give up proper photo face-tagging, FaceTagger is the answer you've been waiting for. Whether you're migrating a decade of family photos or starting fresh, it's the closest thing to "what Photo Gallery should have grown into" — and unlike Photo Gallery, it actually runs on the OS you're using now.
DOWNLOADS:
Windows: (ppApp only for now):
Linux: (LLApp only for now):
Please if you use this and it works or doesn't let me know so I can fix any issues, just reply here and say how it goes. I'll do a full on release soon, but I am pretty worn out polishing the last few bugs getting windows to work as well as Linux was.