Local AI — nothing uploaded
One-time purchase

Find every face in your catalog,
automatically

A Lightroom Classic plugin that recognizes the people in your photos and tags them as keywords. All processing happens on your machine — nothing leaves your computer.

Get the Plugin $19.99
Adobe Lightroom Classic Lightroom Classic
Face Tagger People & Keywords
Features

Built for photographers
who shoot real life

Local AI, Zero Cloud

Face detection and recognition run entirely on your Mac using InsightFace. No photos are uploaded, no servers touched.

Face + Body Matching

Catches people even when their face isn't visible. Uses body detection plus temporal context from surrounding photos.

Automatic Keyword Tagging

Recognized people are added as Lightroom keywords under a "People" parent. Your existing keyword workflow keeps working.

RAW-Aware

Reads .ARW, .CR2, .NEF and other camera RAW formats natively. No pre-export required.

Smart Re-scan

Scans skip photos you've already processed. Re-train when you add a new person and only the unclassified photos are revisited.

One-Time Purchase

$19.99 once. Activate on up to two machines (e.g., laptop + desktop). No subscriptions, no usage limits.

How It Works

Three steps to a tagged catalog

Teach it who's who, then let it do the work.

01

Train

Select one or two photos of each person you want recognized. Name them. That's your whole training set.

02

Scan

Pick a collection, run Deep Scan. The plugin processes every photo locally and matches faces against your trained people.

03

Review

Confirm suggestions for unrecognized faces. Sync corrections auto-tunes matching so the next scan is sharper.

FAQ

Frequently asked questions

What to expect before you install.

Does it really run fully offline?

Yes — after a one-time model download at first run (~500 MB, 2–5 minutes). After that, no part of the plugin needs the internet. Your photos and face data never leave your machine.

Is it Mac-only?

For v1, yes. The bundled AI runtime is built for macOS. Windows users can still run the plugin from source (we've tested it), but we don't ship a Windows installer yet. Tell us if you need Windows — demand drives priority.

How much disk space does it use?

Roughly 500 MB for the InsightFace and YOLO models. Face encodings for your catalog are tiny — about 2 KB per photo, stored in a local SQLite database next to the plugin folder.

Does it support RAW files?

Yes. The plugin reads .ARW (Sony), .CR2/.CR3 (Canon), .NEF (Nikon), .RAF (Fuji), .DNG, and other common RAW formats directly. No need to export to JPEG first.

How many machines can I activate on?

Two. Your license key activates on up to two computers (e.g., laptop + desktop) simultaneously. Deactivate from the Lightroom menu if you switch machines and need a third.

Does it work with multiple Lightroom catalogs?

Yes. Face training and scan history are stored per catalog, so you can use the plugin across unrelated catalogs without them interfering.

Where is my face data stored?

Locally, in face_data.db next to the plugin folder on your Mac. Delete that file (or uninstall the plugin) to remove everything. See our Privacy Policy for the biometric-data details.

How accurate is it?

It depends heavily on your photos. Clear, well-lit faces match reliably after you've trained a person on just one or two reference photos. Side-profile, low-light, partially occluded, and body-only matches (people shot from behind) are inherently less confident and often end up in the review queue rather than auto-tagged.

The plugin is designed around the idea that you'll correct it as you go. After each scan, you confirm or reject suggestions in the review step. Sync Corrections then auto-tunes per-person tolerance based on that feedback. After a few scan + review + sync cycles on the same catalog, accuracy typically climbs into the 90–95% range for the people you've trained — and keeps getting sharper with every round.

What if macOS says "unidentified developer"?

The bundled AI binary is ad-hoc signed. On first launch, macOS Gatekeeper may block it. Right-click the binary and choose Open once — macOS remembers the choice after that.

Can I get a refund?

Yes. Refund requests are handled through LemonSqueezy within the standard refund window. Use our contact form if you run into issues. Full terms on the Terms of Service page.

What if I have questions or problems?

The plugin has a built-in Report a Problem menu item that captures the recent server log and opens our feedback form. You can also reach out directly via the contact form. Solo developer, personal responses.

Stop scrolling looking for "the one with Mom"

Tag your catalog once. Search by name forever.

Get Face Tagger $19.99
Changelog

Release history

1.0.0 April 2026

Initial release

  • Local face detection and recognition (InsightFace + YOLO)
  • Deep Scan with face, body, and temporal matching
  • Train People from Lightroom selections
  • Suggest People clustering for untrained faces
  • Review Unrecognized auto-pass after every scan
  • Sync Corrections with per-person tolerance auto-tune
  • Scan-result collections (per-person, unrecognized, no-face)
  • RAW support (.ARW, .CR2, .NEF, .RAF, .DNG)
  • Built-in update check and Report-a-Problem reporter
  • macOS bundle with PyInstaller sidecar, two-machine license activation