At MAX in October 2025, Adobe shipped Assisted Culling in Lightroom Classic. If you already pay for Lightroom, AI culling help now costs you nothing extra — and that changes the question every dedicated culling tool has to answer. Here’s an honest look at what it does, what Adobe itself says about its limits, and when a dedicated tool is still worth paying for.

One thing this article is not: a benchmark. We haven’t yet run our own timed, like-for-like test against Assisted Culling, so you won’t find made-up numbers here. Where we cite figures, they come from Adobe’s documentation or published third-party tests, credited as such. We plan to publish measured comparisons on the same image sets in future.

What does Assisted Culling actually do?

Assisted Culling analyses your photos locally on your machine and filters them against technical criteria you choose: whether the subject is in focus, whether eyes are in focus, whether eyes are open, and whether there are exposure problems. Criteria can be toggled individually, several have sensitivity sliders, and it can run at import so the filtering is done by the time you start reviewing. Adobe’s documentation has the full walkthrough.

Two things about the design deserve genuine credit. First, the analysis runs on your machine — your photos aren’t uploaded for culling, which matters for the reasons covered in our piece on where client photos go. Second, the criteria are transparent toggles rather than a mysterious verdict: you can see exactly which filters you asked for.

Does it cost extra?

No. Assisted Culling ships inside existing Lightroom Classic subscriptions — there’s no add-on fee and no separate tier. If you’re already paying for the Photography Plan, this is the correct starting point: try the built-in feature before paying anyone else for a second subscription, including us.

That’s not a grudging concession; it’s the honest maths. A dedicated culling tool is another monthly cost on top of Lightroom, so it has to save you enough time — or catch enough frames — to justify existing. For some photographers it won’t. For others, the gap is the difference between a Sunday and an hour, and the sections below are about telling which camp you’re in.

What are its limits right now?

Adobe labels Assisted Culling Early Access, and its own release notes pitch the feature at portraits and headshots; early third-party reviews put that down to training centred on single-subject portraits. Weddings and events, with their mixed lighting, movement and multi-subject scenes, are exactly the territory Adobe’s framing is most cautious about.

Published third-party testing so far points the same way. One head-to-head test — published by Excire, who make a competing Lightroom plugin, so read it with that in mind — clocked the initial analysis at just over 11 minutes on an 879-image portrait set and nearly half an hour on a 1,986-image wildlife set, and found it rejecting tack-sharp images as out of focus — most severely on the wildlife set — while labelling some clearly closed eyes as open. That’s one test, on one machine, on its authors’ own image sets; your results will differ. But “sharp frames falsely rejected” is the kind of miss that forces you to re-check the rejects, which is where the promised time savings quietly evaporate.

In fairness, Adobe is actively improving it — updates since launch have significantly improved how it handles shallow depth-of-field, so intentional background blur is less likely to be rejected as a focus miss. Early Access means exactly what it says: the feature is real, and it’s still becoming what it will be.

Two limits are structural rather than temporary, though:

It filters; it doesn’t rank. Assisted Culling sorts frames against pass/fail technical criteria. It doesn’t grade every frame against the others, and it has no opinion on composition, expression or moment — the dimensions that actually decide which of six sharp frames makes the gallery.
It doesn’t learn your taste. The sensitivity sliders are yours to set, but the feature doesn’t study which frames you pick and adjust to match. Every shoot starts from the same generic criteria.

Who is it genuinely enough for?

Quite a lot of photographers, honestly:

Portrait and headshot shooters — this is the exact use case Adobe says it’s built for, and technical filters (focus, eyes) map well to how portrait sessions fail.
Lower-volume shooters — if a session is 200–400 frames, the cull was never your bottleneck; removing the obvious rejects for free is all the help you need.
Photographers who live entirely in Lightroom — no new app, no new subscription, no moving files between tools.
Anyone whose cull is really “delete the misses” — if you don’t need ranking, grouping, or a picked-for-you first selection, the built-in feature covers it.

If that’s you, use it. It costs nothing, it runs locally, and it will only get better.

When does a dedicated culling tool earn its keep?

The case for a dedicated tool starts where filtering stops being the job:

Volume. A wedding is 2,000–4,000 frames of near-duplicates. The work isn’t finding the blurry ones — it’s choosing the best frame of each moment, which needs burst grouping and ranking, not pass/fail filters.
Event conditions. Dark dance floors, mixed light, movement, groups — the situations Adobe’s portrait-first guidance is most cautious about are the ones an event shooter lives in.
Your taste, not generic criteria. A tool that learns from your past picks starts every wedding already knowing you prefer the laugh over the posed smile.
Explainable scoring. Per-dimension scores (focus, eyes, composition, emotion, moment…) let you trust the picks at a glance and see exactly why when you disagree — a filter can only tell you a frame passed.
What happens after the cull. If the pick list flows straight into a first-pass edit and a client gallery, the time saving compounds; if the cull dead-ends into an export, you’ve moved the bottleneck, not removed it.

This is where LightVision sits: fully on-device culling that grades every frame 0–100 across seven visible dimensions, groups bursts into single decisions, learns your taste from your own picks, and hands the keepers straight to editing, galleries and print sales — from $19 a month. But the honest summary cuts both ways: if you shoot portraits at modest volume inside Lightroom, Assisted Culling may genuinely be all you need, and we’d rather say so than have you pay for a second tool you won’t feel. The full feature-by-feature picture is on our LightVision vs Lightroom comparison page.

The short version

Try Assisted Culling first if you already pay for Lightroom — it’s included, it runs locally, and for portraits and obvious-reject removal it may be enough.
It’s Early Access, portrait-first, and filter-based — that’s Adobe’s own framing; published tests so far echo it, especially for event work.
A dedicated tool earns its keep on volume and judgement — burst grouping, taste learning, explainable ranking, and a cull that flows into delivery.
Distrust anyone’s numbers without a method — including ours, when we publish them. Same image set, same machine, stated dates, or it’s marketing.

Lightroom's Assisted Culling: when it's enough — and when it isn't