How Good Is Lightroom Classic’s New AI Culling? Testing Against Excire Search 2026

AI-powered culling software is not new. If you want to streamline your culling workflow with AI, you have a variety of standalone programs and platforms available: Imagen AI, Narrative Select, Excire Foto 2025, and Aftershoot, to name a few.

But for dedicated Lightroom Classic users, none of these options are ideal; they exist outside of the LrC interface and require extra steps to integrate neatly into a Lightroom workflow.

Up until recently, that was the whole story. Lightroom Classic users were essentially stuck choosing between a variety of bad options. 

And then…lightning struck. In October 2025, Adobe launched AI-assisted culling for Lightroom Classic, a built-in feature designed to streamline photographers’ workflows by automatically selecting the best photos from a batch and rejecting the worst. At the time of writing (November 2025), this feature is still in Early Access. But it’s available to all Lightroom Classic subscribers for free, which makes it an obvious choice for Lightroom Classic users who want an easy way to spend less time on culling workflows. 

Additionally, Excire’s latest Lightroom Classic plugin, Excire Search 2026, was launched several months prior to Adobe’s release. Excire Search 2026 brings advanced AI-assisted culling directly into Lightroom Classic—though you do need to pay for a lifetime license, which gets you unlimited access to Excire culling tools (in addition to various AI-powered search and organization features).

This begs the question: Is Lightroom Classic’s new Assisted Culling feature sufficient for most photographers? Or are Excire’s AI-powered culling tools worth the additional investment?

To find out, I ran a direct comparison. I used Lightroom’s and Excire’s culling tools on two real sets of photos—a portrait session and a bird photography outing—and evaluated the results. Below, I share what I discovered, including comparisons of speed, user experience, control, and accuracy. And I ultimately reveal which AI culling program is best in 2025 (and why). 

Let’s dive right in.

Culling in Lightroom Classic vs Excire Search: The Comparison

Lightroom Classic needs little introduction. It’s been an industry-standard photo organizer and editing program for well over a decade, and I’ve personally used it for the better part of that period. 

However, this comparison marks my first real test of Lightroom Classic’s brand-new AI-assisted culling tools. Since the feature only recently debuted, I’m approaching it with fresh eyes and no preconceived notions about its effectiveness.

Excire Search 2026, on the other hand, is a Lightroom Classic plugin that offers a variety of AI-powered tools for photo-management, including automatic keywording, AI prompt search, duplicate detection, and smart culling. I’ve been using Excire Search 2026 inside Lightroom for a few months now, so I’m more familiar with its approach to culling. To avoid any bias, though, I tested both tools on the exact same sets of images and evaluated their selections from a fresh, side-by-side perspective.

Now, both systems claim to identify the “best shots” automatically, and they both do it non-destructively and without ever leaving Lightroom, but they differ significantly in approach. Importantly, both tools also process images entirely on your local machine. That means there’s no cloud upload, no external server involvement, and no need to worry about sensitive data being sent off your computer.

In short, the comparison ahead focuses on how these two AI-powered photography tools perform when asked to do the same job: find the best photos quickly, accurately, and reliably.

How I Tested Lightroom and Excire

Note: If you’re eager to jump straight to the verdict, feel free to skip ahead to the results section below. But for those who want a clear picture of the methodology, here’s the full breakdown. 

To compare Lightroom Classic’s new Assisted Culling with Excire Search 2026’s dedicated culling tools, I used real-world photoshoots in two very different genres:

1. Portraiture: Here, I used a couple’s pre-engagement shoot from a professional portrait photographer; the batch contained 879 images.
2. Wildlife/Bird photography: I captured these 1986 shorebird and wading bird photos on a fun visit to the Florida coast back in 2018. 

In both cases, the goal was to narrow the set down to a small collection worth editing.

Now, Lightroom’s culling is currently optimized for portrait images, so I didn’t expect it to perform strongly on wildlife. But Excire’s AI culling isn’t explicitly portrait-centric, so I thought it valuable to observe how both systems handle non-human subjects in a traditional wildlife photography scenario (thousands of photos, many of them blurry, most of them shot in bursts).

The Lightroom Classic Setup

For both shoots, I imported the images normally into Lightroom Classic and then started the analysis by opening the Assisted Culling panel in the Library module. While Lightroom ran its analysis, I chose my selection criteria. 

For the portrait session:

• I set the Eye Focus slider to 100 (fully toward “Sharp”), which told Lightroom to select only the images with maximally sharp eyes
• I checked the Eyes Open box, which told Lightroom to reject images where eyes were closed; I also checked the Reject Photos Without People’s Eyes option.

I also told Lightroom to:

• Reject Documents and Receipts
• Reject Misfires
• Reject Exposure Issues

For the bird/wildlife shoot, the process was similar. I initially hoped to use Eye Focus for birds, but after checking a variety of photos, it was clear that Lightroom’s culling system didn’t detect bird eyes at all. Instead, I set the Subject Focus slider to 100. I used the same set of Reject settings as in the portrait test (Documents and Receipts, Misfires, Exposure Issues). 

In both cases, once the culling process was complete, I used the Batch Actions option to apply green color labels to Lightroom’s selections and red color labels to its rejections, which made it easy to evaluate the results.

The Excire Search 2026 Setup

To use the Excire Search plugin, you must first initialize your images in Lightroom Classic. During this process, Excire goes through your files and makes them searchable with prompt search, facial recognition, and more; initialization is also when Excire adds descriptive keywords to images.

So I started with initialization, then I created a new culling project and dialed in some basic settings. Excire offers quite a few more culling options than Lightroom—for instance, you can tell Excire to break down your image batches into groups based on content, people, capture-date/time, visual similarity, and more. And you can tell the AI to select your images based on sharpness, face sharpness, eye sharpness, eyes open, smile, and aesthetics.

For the portrait session, I grouped images by visual similarity, chose Content (rather than Color) as the comparison feature, and left the Strictness slider at its default value of 25. (My goal was to create small sets of photos similar enough to compare meaningfully without becoming too broad or too fragmented.)

I then instructed Excire to automatically reject:

• Blurry photos
• Overexposed and underexposed photos
• Photos in which all eyes were closed
  

Then in the Smart Selection tab, I told Excire to automatically select the top two images from each visually similar group, using the criteria Eye Sharpness, Eyes Open, and Aesthetics.

For the wildlife shoot, I again grouped by visual similarity (Strictness: 25) using Content as the comparison feature. I instructed Excire to automatically reject:

• Blurry photos
• Overexposed or underexposed photos

In the Smart Selection tab, I instructed Excire to choose the top two images from each group using Sharpness and Aesthetics as the criteria.

Evaluating the Culling Tools

Once the culling tests were complete, I evaluated the workflow and results in four key categories:

• Speed: How long did each system take to cull the images?  
• User Experience: How easy and intuitive was each tool to use?
• Control & Flexibility: How much influence did I have over the selection process, and how adjustable were the settings?
• Accuracy: Did each system choose the same “keepers” I would have selected manually?

For each category, I scored Lightroom Classic and Excire Search 2026 on a scale from 1 to 5, where 5 represented exceptional performance and 1 indicated a deeply disappointing result.

Excire Search 2026 vs Lightroom Classic: AI Culling Test Results

Before getting into the specifics, it’s worth noting that both Lightroom Classic and Excire Search 2026 handled the test images without technical issues—always a good start! The real differences emerged in how quickly the tools processed large batches, how intuitive they felt to use, how much control they offered, and how accurately they identified the strongest shots.

Speed

One of the biggest concerns with AI-assisted culling is the time required for analysis. If you’re working under tight deadlines—or simply dislike waiting around—you don’t want to spend hours watching a progress bar before you can start reviewing your selects. So for this comparison, I made sure to carefully measure how long each system took to analyze my test batches.

Although Lightroom Classic offers the option to run its Assisted Culling during import, I chose to begin the process manually for both programs so I could measure the analysis time accurately. It’s also important to note that analysis speed varies depending on the number of photos and the power of your computer. My test batches were fairly large and my hardware is modest, so I expected both programs to take a while.

What mattered most to me, however, is the relative performance: how Lightroom Classic compared to Excire Search 2026 when given the same task.

Lightroom Classic Analysis Times

• Portrait batch (879 images): 11:22
• Wildlife batch (1,986 images): 29:30

Excire Search 2026 Analysis Times

Excire’s workflow typically involves two stages. First, you initialize your images, which makes them searchable via Excire’s full toolkit and automatically adds descriptive keywords. Second, you run a dedicated culling project, where Excire can group and select/reject images based on your criteria.

Below, I’ve listed both times for each batch:

Portrait batch (879 images):

• Initialization: 6:02
• Culling: 3:44

Wildlife batch (1,986 images):

• Initialization: 12:55
• Culling: 7:45

Neither program was painfully slow, but Excire worked at a noticeably faster pace—especially once images were initialized. And if you already use Excire for photo search and organization, the culling step itself will feel especially fast. Lightroom Classic took significantly longer, especially on the wildlife batch; nearly 30 minutes is manageable, especially if you don’t mind leaving the tool to run while you work on something else, but it’s not ideal if you routinely process large sets of images.

Lightroom Classic speed rating: 3/5
Excire Search 2026 speed rating: 4.5/5

User Experience

A culling tool may offer clever AI features, but if it’s difficult or unintuitive to use, it quickly becomes more trouble than it’s worth. With that in mind, here’s how Lightroom Classic and Excire Search 2026 stacked up in terms of day-to-day usability:

If you’re already a longtime Lightroom Classic user (as I am), Adobe’s new Assisted Culling tool will feel easy and familiar. You simply open the Assisted Culling panel in the Library module and Lightroom immediately begins analyzing your images. While the analysis runs, you can set your culling preferences, and the interface presents these options in a clean, straightforward way.

Landing on the right settings for your workflow is also fairly simple, in part because Lightroom doesn’t offer many choices. The only slightly tricky element is determining where to position the Subject Sharpness and Eye Sharpness sliders. In practice, I found that pushing these sliders to 100 was generally the best approach for portrait photos; any lower, and Lightroom returned far too many selects to be helpful. (For example, during the portrait culling test, lowering the Eye Focus slider to the 90+ range produced 740 selects out of 879 images—far too many to be of use.)

Once culling is complete, Lightroom’s presentation of results is solid: Selects are marked with green icons, rejects are marked with red icons, and you can easily toggle between viewing all images, only selects, or only rejects. The Batch Actions button allows you to automatically apply flags, color labels, or star ratings, and you can choose to send selects or rejects directly into a Collection for faster organization.

Getting started with Excire’s culling tools is also straightforward. In the Excire Search panel, you click the Start/Adjust Culling Project button and configure your preferred settings in the dialog that appears. The workflow is slightly more complex than Lightroom’s, primarily because Excire offers more sophisticated options.

Excire’s culling offers three main culling layers:

• Grouping: Excire can break your photos into meaningful groups—what I like to think of as photoshoot “chunks”—based on visual similarity, content, people/faces, capture date/time, or sequence order.
• Sorting: Excire can evaluate a group using adjustable criteria and sort (i.e., order) the images accordingly. 
• Smart Selection: Here, Excire evaluates each group using adjustable criteria and automatically flags (or stars, or color-labels) the top images within each group.

Importantly, Excire never forces you to let the AI choose your images. If you prefer, you can instruct it to group the files—for example, by visual similarity—and then manually review each group yourself (with help from Excire’s Sorting option as needed). 

And these groups are easy to work with; each automatically appears as a Collection in Lightroom Classic. If you enable Smart Selection, you’ll also get a dedicated Selection Collection, which contains all of Excire’s top picks in one place.

(Excire also includes a Rejections tab, where you choose criteria that should trigger automatic rejection. Any images rejected by the AI are added to Excire’s Recycle Bin Collection, making it easy to review or delete them if needed.)

All of this may sound like a lot to configure, and when I first started with Excire Search 2026, it did take me a few minutes to fully grasp the culling workflow. But after one or two uses, the process became very natural. I’d start a culling project, choose my grouping options, set my Smart Selection criteria, and let Excire handle the rest. Once the analysis was finished, I found all my best photos neatly gathered in the Selection Collection, ready for editing.

Excire also provides several helpful tools for manual review, which you can access in the Excire Search Panel. The Culling Group view gives you an overview of all grouped sets:

And when you browse images containing people, Excire automatically displays a magnified face preview and applies color labels based on face sharpness:

The latter feature wasn’t helpful for the bird photos, of course, but it made my final pass through the portrait selects much easier.

So which AI culling tool proved easier to use? Overall, I think Lightroom Classic has a slight edge. Its culling workflow includes fewer options, which makes it faster to understand. That said, once you’ve used Excire’s culling tools a few times, the process becomes nearly automatic—and the dedicated Selection Collection is a highly convenient feature that Lightroom doesn’t offer.

Lightroom Classic user experience rating: 4.5/5
Excire Search user experience rating: 4/5

Control & Flexibility

This is an area where Excire Search 2026 really shines. Excire’s culling features are highly configurable, offering a wide range of options that let you tailor the process to your own preferences and shooting style. When you cull with Excire, you’re not boxed into a fixed workflow; instead, the plugin adapts to the way you like to work.

When setting up a culling project, you can instruct Excire to group your files based on people, content, visual similarity, capture date, or sequence order—whichever makes the most sense for your shoot. You can even create multiple types of groups for the same batch of images.

In practice, I tend to use Visual Similarity as my core grouping method across most genres—portraiture, wildlife, landscapes, street photography, and more—because it neatly clusters redundant images. For weddings and other events, however, the People grouping becomes especially helpful, and when I want to get a sense of what went on throughout a shoot, the Capture Date grouping (set to quarter- or half-hour intervals) is great. I recommend that you spend a bit of time experimenting to see which groups work best for you; for instance, while I prefer the Visual Similarity option for wildlife and bird photography, other photographers may find the sequence order grouping more effective (it groups bursts of images).

You can also decide whether you want Excire to automatically choose the best images from each group. And here, you can specify exactly which criteria Excire should prioritize: general sharpness, aesthetics, eye sharpness, face sharpness, smiles, and more. You can even tell Excire how many images to select from each group (or, alternatively, what percentage of each group should be selected). This gives you fine-grained control over how many final picks you receive.

This is fundamentally different from Lightroom’s Assisted Culling. With Lightroom, you have no meaningful grouping options aside from the Auto-Stack feature, which is meant to cluster similar images but is tough to integrate into a culling workflow. More importantly, Lightroom gives you no control over the number of images it selects. You are essentially forced to accept whatever quantity the AI delivers—even if that means sorting through hundreds of so-called “best” shots in order to identify your true keepers. 

Lightroom also offers only three variables for determining what counts as a “best” image: Subject Focus, Eye Focus, and Eyes Open. Excire includes these options, too—but it also offers additional criteria, including aesthetics, face sharpness, and whether the subject is smiling.

I also want to make clear that Lightroom Classic’s Assisted Culling is explicitly designed for photos of people. And at the time of writing, with the tool in the Early Access phase, Adobe has noted that it’s not yet good enough to be used for event photography. This severely limits its potential, and it’s unclear whether Adobe ultimately plans to expand its usefulness beyond portraits and basic event work.

Contrast that with Excire’s culling tools, which certainly offer people-specific features but also work across a wide range of genres, including wildlife, landscapes, architecture, macro, sports, and travel. That breadth of applicability, combined with the plugin’s extensive customization options, makes Excire dramatically more flexible in real-world workflows.

Lightroom Classic control & flexibility rating: 2/5
Excire Search 2026 control & flexibility rating: 5/5

Accuracy

Above all else, accuracy is what matters with AI culling. A speedy, easy-to-use culling tool that can’t recognize a blurry photo is essentially useless. With that in mind, I was curious to see how Lightroom Classic and Excire Search 2026 would stack up in practice.

Lightroom Classic

For the portrait photoshoot cull, Lightroom selected 497 images and rejected 382.

That already felt like far too many selects—roughly half of the entire shoot. About 50% of the images were marked as having 100/100 Eye Focus, which meant I now had to sift through several hundred photos to find my true keepers—the very task the AI is supposed to reduce.

And when I did go through the individual images, I ran into additional problems. Many photos were rated 100/100 for both subject sharpness and eye sharpness, yet they weren’t all that sharp. This next image is one that Lightroom selected, yet the main subjects are clearly out of focus (for better viewing, I’ve zoomed in to 100%):

And when I looked at the Culling Scores panel, which allows you to view the scores Lightroom has attached to each photo, the ratings were often confusing. Several images had 100/100 eye sharpness but noticeably lower subject sharpness scores, which doesn’t reflect real-world photographic logic: if the subject isn’t sharp, the eyes shouldn’t be sharp, either.

Lightroom’s eye-state detection was also inconsistent. Some images with clearly closed eyes were labeled “Likely open,” while others with open eyes were marked “Likely closed.”

To Lightroom’s credit, it did a decent job identifying blurry images caused by camera shake. Several obviously unsharp photos received appropriately low sharpness ratings, like this one here:

But overall, the inconsistent evaluations and overwhelming number of selects meant that Lightroom’s Assisted Culling simply wasn’t reliable or sensitive enough to genuinely speed up my workflow.

What about the wildlife photography cull? As expected—given that Adobe explicitly states the tool is designed for people—performance on bird photos was poor.

This time, the results were the reverse of the portrait test: Lightroom selected only 60 out of 1,986 bird images. Because the AI couldn’t detect bird eyes, I relied solely on Subject Focus, but even then Lightroom incorrectly rejected a huge number of tack-sharp photos, such as this one here:

At the same time, several of the few images rated 100/100 for subject sharpness were visibly blurry:

Had I been culling this wildlife set for publication on my website or social media, I would have ignored Lightroom’s assessment entirely and culled manually.

Excire Search 2026

Since Excire allows you to specify the number of images selected per group, I ended up with 138 selects from my portrait cull—two images from each visually similar group. This produced a much smaller, much more manageable set of candidates.

But were they actually good?

I was honestly impressed. At best, the images in my Selection Collection were outstanding; at worst, they were technically sharp and usable. In some cases, I disagreed with Excire’s top choices, but this usually came down to subjective preference. Overall, I was pleased: Excire had narrowed the portrait session down to a high-quality batch from which I could quickly choose my final favorites. And had I instructed Excire to choose only one image per group, I would have ended up with an even tighter set.

And when reviewing the bird photoshoot, I was once again surprised by the quality of the picks. Excire doesn’t offer wildlife-specific features, yet its visual similarity groupings were almost flawless. The plugin divided my nearly 2,000 images into about 100 groups and selected the top two photos from each based on sharpness and aesthetics.

Overall, I liked the picks. A few blurry images did appear in the final selections, but that wasn’t really Excire’s fault; when photographing birds in low light, I often end up with at least a few sequences of motion-blurred shots, so in some cases the AI simply had to choose the “best” among a slew of technically flawed frames.

Even more surprising was how well Excire handled my more artistic bird photos. When I originally took them, I was experimenting heavily with backlighting and sun flare, but Excire still managed to identify strong images in each group.

If I have one small criticism, it’s that Excire didn’t seem to consider bird head angle as part of its aesthetic evaluation. When I’m photographing birds, I am pretty picky about this—but it’s a minor complaint, especially given that Excire isn’t specifically designed for bird workflows.

In the end, Excire genuinely shaved time off my wildlife culling workflow. By grouping the images effectively and producing a high-quality selection set, it gave me a great starting point for a quick final pass.

Lightroom Classic accuracy rating: 2.5/5
Excire Search 2026 accuracy rating: 4/5

Overall Verdict

Lightroom Classic’s Assisted Culling tool does have strengths. It’s smoothly integrated into the Library module, and because it lives within the Lightroom environment, it feels immediately familiar and easy to use. In my testing, I never struggled with the interface—though the analysis speed was on the slow side, especially for large batches.

Where Lightroom fell short was in flexibility and, more importantly, accuracy. The tool simply doesn’t feel ready. It doesn’t offer enough meaningful culling criteria, and even within the limited set of options Adobe currently provides, the results were inconsistent. For portrait work, Lightroom’s sensitivity to sharpness was far too low; even with sliders set to 100, the AI marked hundreds of images as “best” despite obvious technical flaws. And while Lightroom did manage to identify extremely blurry frames, too many remaining images were given perfect or near-perfect scores to make the selections useful—and there was no practical way to refine or reduce them further. Ultimately, AI culling only works if you trust the software to evaluate your photos effectively. But after using Lightroom’s Assisted Culling tool, I don’t trust the AI to make useful, reasonable decisions.

Excire Search 2026, on the other hand, offered a far more powerful and comprehensive AI culling experience. Its people-focused tools were genuinely helpful in the portrait session, and for wildlife, the general sharpness and aesthetics criteria performed surprisingly well. (I’ve also used Excire Search to cull landscape, architectural, and street photography batches and had similarly positive results.)

With Excire, I was able to reduce huge collections of images—sets that would normally take hours to review—down to small, highly usable selections. In real-world workflows, that kind of reduction saves a lot of time.

The flexibility is also a major advantage. Excire offers a wide range of grouping options that can be tailored to different genres and shooting styles, and the various culling criteria can be mixed, matched, and experimented with as needed. And in situations where I don’t want the AI to choose the best images for me—such as when I’m working on more experimental fine-art projects—the ability to simply group the images by visual similarity and review them manually is invaluable. Excire essentially gives structure without taking control away.

So which culling tool works best for a Lightroom Classic workflow? For me, Excire Search 2026 is the clear winner. While it does require a paid license, there is no ongoing subscription to worry about. Plus, you can try it free for 14 days by downloading a trial from the Excire website.

Bottom line: If you value accuracy, flexibility, time savings, and trust, Excire is the culling tool that will genuinely streamline your Lightroom Classic workflow.