Everything between field work and a named catalog.

Match across 150,000+ species, on your machine.

Nomen runs a vision model trained on millions of verified specimen photographs and field observations. It covers terrestrial and aquatic life across 29 downloadable packs, from birds and mammals to nudibranchs and corals. Every match returns the top 10 suggestions with confidence scores, and everything runs locally. No cloud, no uploads, no external calls.

When Nomen is not sure enough about a match, it skips the photo rather than guessing. You control how strict it is. No false certainty.

Underwater photographers can select all 15 aquatic packs with one click. Download only what you shoot. Packs install through the built-in download manager and run entirely on your machine.

Works where your photos already live.

Nomen installs as a standard Lightroom Classic plug-in through the Plug-in Manager. There is no separate app to launch, no browser to open, no images to export. Select photos in the Library module, click Match, and results are processed and written to your catalog immediately.

Fast enough to keep up with your shutter.

On Apple Silicon Macs, Nomen uses the GPU and neural engine to run both the detector and the species matcher. A 72-photo batch runs in about 27 seconds on an M4 Mac mini, with full subject detection and bounding boxes on every frame. Roughly 0.4 to 0.5 seconds per photo once the model warms up.

Windows currently runs on CPU. Performance is stable and predictable; GPU acceleration for Windows is the next major item on the roadmap.

Acceleration is automatic. No configuration, no separate install, no model selection. Nomen picks the fastest path available on the machine it is running on.

One subject, every subject, or the whole flock.

Nomen offers four detection modes, so matching fits the photo rather than forcing the photo to fit the tool.

Main subject (recommended) matches the single most prominent species in the frame. Best for photos with a clear subject: a portrait, a perched bird, a lone animal on the reef. One species, one result, written directly to your catalog.

Every subject finds and matches all species in the photo. A reef scene with a moray, a cleaner wrasse, and a nudibranch behind them. A wetland shot with four species of wading bird. Nomen detects each one and presents a multi-species review grid where you accept or reject each match individually. All accepted species are written as separate keyword paths.

Fast survey counts and classifies every individual at standard resolution (10 to 30 seconds per photo). Tiles the image into overlapping sections so distant birds in a flock, individual fish on a reef, or scattered mammals across a landscape all get detected. Shows a per-species count summary with a detection overlay.

Deep survey is the same counting workflow at full sensor resolution (1 to 3 minutes per photo). Use it when subjects are small or distant and Fast survey is missing them. Designed for bird flock photography, underwater reef surveys, wildlife census work, and research documentation. Counts on very small or very distant subjects are approximate.

All four modes write the same structured metadata. The difference is whether you want Nomen to focus, to be thorough, or to count.

Review as much or as little as you want.

Pick a review mode in the Run dialog and Nomen tunes how many eyes are on the results.

Don’t review, apply all matches is straight-through: no dialogs, results written as they complete. Best for batches you trust.

Quick review (grid) opens a three-column grid at the end of the batch with a thumbnail and a checkbox per photo. Uncheck anything that looks wrong; the rest are applied in one go. The grid title shows total batch time so you can see how long the run actually took.

Full review (one photo at a time) pauses on every match, shows the bounding box plus the top species and ranked alternatives with rarity and confidence labels, and lets you confirm or pick a different suggestion. Close-match reviews share the same dialog, so the layout is consistent whether Nomen is pausing because the call is tight or because you asked it to pause on every photo.

Names, keywords, captions, and searchable fields, written consistently every time.

• Captions: Common Name (Scientific Name) with a confidence label, plus social-ready captions with EXIF data and auto-generated hashtags.
• Keyword hierarchies: a Taxonomy tree (Nomen > Class > Family > Genus > Species), plus optional Region and Common Names branches. Multi-species photos get every species as its own keyword path.
• Multilingual keywords: 12 languages to pick from (English, French, German, Italian, Japanese, Dutch, Norwegian, Polish, Portuguese, Spanish, Swedish, Chinese), with a separate display language for the review UI.
• Star ratings: derived from image quality scores.
• Color labels: set by confidence level (green, blue, yellow, or red).
• Custom metadata fields: scientific name, common name, confidence, family group, match status, detection label, rarity label, and region used.

This is not just tagging. It is catalog infrastructure that stays useful for years.

Social-ready captions, written for you.

Nomen writes captions formatted for sharing. It pulls the species name, location, camera model, focal length, aperture, shutter speed, and ISO from your photo’s EXIF data, then generates hashtags from the species name and taxonomic group. The result is a caption ready to copy and post.

A blank line is left for your notes. Add your story, the encounter, or nothing at all. You were there. The caption just saves you the formatting.

When it is not sure, it asks you.

When the top two suggestions are too close to call, Nomen asks you to choose. Keep the top match, pick another from the list, or mark the photo as unmatched. You set how cautious Nomen should be. More cautious means more review dialogs, less cautious means more automation.

Pre-match summary lets you see exactly what will be written before anything touches your catalog. Multi-species review presents a grid of detected subjects with checkboxes to accept or reject each one individually, so you stay in control when Every subject mode finds more than you expected.

Every suggestion includes a one-click verification link to an external species reference, so you can double-check without leaving the review flow. And if a match turns out wrong after the fact, native Cmd-Z (macOS) or Ctrl-Z (Windows) reverts the whole batch as a single Lightroom undo, including keywords, captions, ratings, and labels. Nomen keeps 10 rollback snapshots so you can always step back.

No silent wrong answers. No pretending certainty where it does not exist.

A dozen photos or a thousand. Same workflow.

Select a morning’s shoot or an entire field trip. Nomen processes them all locally and writes results to your catalog as they complete. Photos you have already matched are automatically skipped. Manual IDs you have made are preserved.

Use Apply Latest Results to re-apply previous match data without re-running the model, or export results to CSV when you need them outside Lightroom.

Geography sharpens the match.

When your photos carry GPS coordinates, Nomen cross-references species occurrence data from 203 countries. A bird photographed in Costa Rica will not be mismatched as a lookalike found only in Southeast Asia. A Snow Bunting in Ontario will not be confused with McKay’s Bunting.

Frequency-weighted scoring means common local species rank above rare lookalikes, and the match reflects what you are actually likely to have photographed. Each suggestion is tagged with a rarity label (very common, common, uncommon, rare, not recorded) based on local observation frequency, so you can judge a result at a glance.

Underwater photographers can select dive regions (Coral Triangle, Red Sea, Caribbean, and others) to narrow matches to the species they are most likely to encounter. If no strong match is found in your country, Nomen offers to search neighboring countries before giving up.

Choose from 29 species packs through the built-in download manager. Geographic filtering is automatic when GPS data is present and gracefully degrades when it is not. The model still runs, just without the geographic filter.

Your photos never leave your computer.

The model runs entirely on your machine. No cloud servers. No uploads. No external API calls during matching. The only internet connectivity required is a one-time download of species packages and periodic license validation every 30 days, with a 30-day grace period for offline use.

For photographers who care about privacy, speed, or working in places without reliable internet, local is not a tradeoff. It is the point.