LightVision Studio vs Imagen
Imagen is a mature AI editing service with a deep Lightroom-plugin workflow: it learns your style and returns real develop settings to your catalog, processing your photos in the cloud. LightVision keeps the whole job — culling, style-matched editing, face AI, even generative fill and extend — on your Mac, and adds galleries and print sales.
Last reviewed July 2026. Imagen's features change — verify current details on their site.
| Capability | LightVision | Imagen |
|---|---|---|
| AI culling with quality scoring | AI culling, 7-point grading | Yes — grouping, ratings, blink/focus flags |
| Explainable per-photo dimension scores | Yes — inspectable per-photo scores | Grouping + ratings |
| Auto first-pass edit in your style | Yes — learns from your finished work | Yes — cloud profiles |
| Separate style profile per shoot type | Yes — separate per shoot type | Yes — multiple profiles |
| Deep Lightroom-catalog integration | No — own editor; finish anywhere | Yes — its home turf |
| Platforms | Mac (Apple Silicon recommended) | macOS & Windows |
| Where AI processing runs | On your Mac, on-device by default | In the cloud (AWS, per Imagen’s docs) |
| Works offline / on location | Yes — no upload step | No — needs internet to process |
| Do originals leave your machine? | No | Yes — auto-deleted within ~7 days; low-res previews retained (their docs) |
| Duplicate & burst grouping | Yes | Yes |
| Auto moment collections + story arc | Yes — on-device, shoot-type aware | No |
| Click-a-face person search | Yes — click a face, identity match | Varies |
| Per-person “Require N in Picks” quota | Yes — require N in Picks per person | No |
| Family / group shot-list coverage | Yes — flags missing group shots | No |
| Opt-in guest face-search in client gallery | Yes — opt-in, anonymous, in-browser | No |
| Borrow open eyes from a burst frame | Yes — borrows open eyes from the burst | No |
| Type-a-word AI masking | Yes — type a word → editable mask | No |
| On-device semantic search | Yes — on-device, no API key | No |
| Client galleries (delivery) | Yes — polished client galleries, free tier | No — exports to gallery services |
| Print store / sell prints | Yes — pro-lab fulfilment, 0% commission on paid | No |
| Sell digitals safely (watermark + gated originals) | Yes — baked-in watermark, gated originals | No |
| Pricing model | Free tier, or $180–$660 / year (flat) | Per-photo ($0.05 PAYG) or volume plans |
What Imagen does well
- A deep, mature Lightroom-plugin workflow — edits come back as real develop settings in your catalog.
- Well-regarded personal editing profiles, plus a large Talent-profile ecosystem to start from a known editor’s look.
- Cloud processing spares your computer — it runs comfortably on modest hardware, and unlimited culling comes bundled with its larger editing plans.
Where LightVision is different
Where do your photos go?
Imagen is upfront about this in its own support docs: photos are processed in the cloud on AWS, originals are automatically deleted within about 7 days, and low-resolution previews are retained to keep tuning your editing profile. That is a disclosed, reasonable design — but it means client work leaves your machine on every job. LightVision runs culling, grading, style learning, face AI and even generative fill and extend on your Mac. Nothing has to be uploaded, so there is nothing to delete later.
Flat monthly vs per-photo: the 5,000-frame wedding
Imagen bills per photo: the published pay-as-you-go rate is $0.05 per AI-edited photo (with a small monthly minimum), extra AI tools such as crop or straighten add $0.01 per photo each, and annual volume plans start around $59 a month. Cull a 5,000-frame wedding (on pay-as-you-go, culling itself is a separate add-on subscription) and edit 1,500 keepers, and that is about $75 per wedding before add-ons — in the region of $2,250 over a 30-wedding year, or Imagen’s largest published volume plan at roughly $199 a month. LightVision is a flat $19–$69 a month with a free tier: the second wedding of the month costs nothing extra. To be fair the other way: if you shoot only a handful of small jobs a year, Imagen’s per-photo model can genuinely be cheaper — flat pricing wins as volume grows.
No upload leg, no queue
A cloud workflow adds a structural step to every job: generate previews, upload, wait for processing, download results. On a hotel connection after a destination wedding, that step is the workflow. LightVision has no upload leg at all — the AI starts the moment the card finishes importing, works on a plane, and is exactly as available in a venue basement as in your studio.
Culling that explains itself
Imagen’s culling groups similar frames, flags blinks and focus misses, and rates your shots — bundled with its larger editing plans, and a separate add-on subscription on pay-as-you-go. LightVision goes further on the decision itself: every frame is graded 0–100 across seven named dimensions — clarity, focus, open eyes, warmth of emotion, composition, energetic moment, exposure — so a Pick is explainable, not just a rating.
One app from import to income
Imagen edits and hands off — it can export finished files or push them to gallery services. LightVision carries the same shoot from cull to a delivered, monetized gallery itself: cinematic slideshows, polished client galleries and a lab-fulfilled print store are built in, with no per-photo fees on any of it.
Auto-built moment folders and a story arc
LightVision sorts each shoot into named moment collections — Getting Ready, Ceremony, First Kiss, First Dance, Cake Cutting, Speeches, Dancing — tuned to the shoot type, then builds a chronological Story Arc that orders your best picks through the day once they’re graded. The moment folders are detected on-device with no API key before any grading, so an import starts organizing itself, not just flagging keepers.
Click a face, and hit a quota per person
On-device face recognition groups people by who they actually are across the whole shoot; right-click a face to see every shot that person is in. The People panel then lets you require a number of shots per person in the final Picks — 1, 2, 3 or 5 — and flags anyone who falls short, with per-person progress like “2 of 3 in Picks · needs 1 more”, biggest gaps first, and one click — “Add best shot of everyone missing” — to pull each short person’s top-ranked unpicked frames in.
Name the bride once — it sticks all shoot
Right-click a face to name a person, or tag a built-in wedding role — Bride, Groom, Couple, Bride’s family, Groom’s family, Best man, Maid of honour, Wedding party or VIP guest. The name binds to that face’s identity, not to a throwaway “Person 7” number, so it re-attaches to the right person by face similarity as the cull re-clusters. Names and roles save into your project file locally — no cloud account or upload — letting you filter and prioritize the whole shoot by who matters.
Did the family-together shot land?
Define must-have group shots from named people — couple plus each set of parents, the whole wedding party — and LightVision checks whether a frame with that whole group together actually made your Picks. It finds the together-frames by intersecting each person’s face matches, counts how many reached Picks against a required count (1–3), and tells you the exact failure: a member isn’t in this shoot, no single frame has everyone together, or together-frames exist but aren’t picked yet. One click adds the best together-frame to fill the quota.
Guests find their own photos by face
On a delivered client gallery you can flip an opt-in toggle — off by default — that lets a wedding guest tap their own face and filter the gallery to just the photos they’re in. It’s private by design: only anonymous face vectors are published (no names, no labels), matching runs entirely in the guest’s own browser with nothing sent to a server, and the face picker only appears after the gallery password gate is unlocked.
Fix a blink without losing the shot
When someone blinks in a burst, LightVision borrows that person’s actual open eyes from a sibling frame and composites them back as a non-destructive layer. In a group photo where two people blinked, each is fixed independently, and every fix is identity-verified by face recognition before it’s applied.
Rescue faces shot in dim light — locally
On-device AI face restoration automatically flags soft or grainy faces — the noise high-ISO receptions and ceremonies produce — and restores their detail with a 0–100 strength slider, blended back non-destructively. It targets every detected face, not just the center subject, and the frames never leave your Mac to do it.
Type a word to make a mask
In the editor, type a region like “the sky” or “her dress” and LightVision builds a pixel-accurate, editable adjustment mask on-device. It’s open-vocabulary — name almost anything in the photo — then dial exposure and color on just that region.
Search your shoot in plain English
Type “kids laughing” or “golden hour by the lake” and on-device AI finds the shots that look like that — matching the meaning of your words to the actual pixels, not keyword tags. It’s the default search, needs no API key, and works on an unculled import out of the box.
Sell digitals without giving them away
Because LightVision delivers and sells in the same app, it also protects what you sell. Previews carry a watermark baked into the pixels (so it survives a screenshot), the full-resolution originals are kept off the gallery page and released only to the client who bought or earned them, and an “included” count like 30 images is enforced per client on the server — clearing cookies won’t reset it. Imagen’s job ends at the edited files, so protected delivery happens in whatever gallery tool you pair it with. (No tool can stop an outright screenshot — but the baked-in watermark rides along on it.)
Is LightVision a good Imagen alternative?
Is LightVision a good Imagen alternative?+
Yes, if you want the work done on your own Mac and a flat bill. LightVision matches the shape of Imagen’s promise — culling plus a first-pass edit in your learned style — but runs everything on-device, including generative fill and extend and face AI, and adds client galleries, slideshows and a print store for a flat $19–$69 a month with a free tier. If your workflow is built around Lightroom catalog round-trips and Talent profiles, Imagen is genuinely strong there — that part is its home turf.
Where do my photos go with each tool?+
Imagen’s own support docs describe its design plainly: photos are processed in the cloud on AWS, originals are automatically deleted within about 7 days, and low-resolution previews are retained to keep tuning your editing profile. That’s disclosed and reasonable — but client work does leave your machine on every job. With LightVision, culling, grading, style learning, face AI and generative fill and extend all run on your Mac: nothing has to be uploaded to edit a wedding.
What does a 5,000-image wedding cost on each?+
Imagen bills per photo: editing is $0.05 per photo pay-as-you-go (culling is a separate add-on subscription there, bundled with its larger plans), and extra AI tools like crop or straighten add $0.01 per photo each. Edit 1,500 keepers from a 5,000-frame wedding and that’s about $75 before add-ons — roughly $2,250 over a 30-wedding year, or its largest published volume plan at about $199 a month. LightVision is flat: $19, $39 or $69 a month whatever your volume, with a free tier. For a handful of small shoots a year Imagen’s per-photo model can be cheaper; at wedding volume, flat pricing wins.
Does LightVision learn my style like an Imagen profile?+
Yes. LightVision reads your real develop settings (exposure, contrast, highlights, shadows, whites, blacks, clarity, vibrance, saturation, temperature, tint) from your finished work, builds a per-shoot-type profile, and applies a matched first-pass edit to new imports — all computed on your Mac, so your style profile never lives on someone else’s server.
Do I need an internet connection to use LightVision?+
Not to cull or edit. Culling, grading, style learning, semantic search, face AI and generative fill and extend run on-device, so the flight home from a destination wedding is working time. You’ll need a connection for the parts that are genuinely online — publishing a client gallery and receiving print orders.
Can I keep finishing in Lightroom?+
Yes. Many photographers cull and first-pass in LightVision, then finish favorites in Lightroom. The difference from Imagen is that the first pass happened locally with no per-photo fee — and LightVision learns from your finished work either way.
See how LightVision compares on your own shoot.
LightVision is in beta — join free, no card required, and we'll help you get set up.