eufy Omni S2 review: obstacle leader, carpet question

The S2 is eufy's flagship vacuum and roller mop. Its dock, a tower 26.4 in tall, empties the bin and washes the mop roller in 140 °F (60 °C) water. The robot navigates with a camera and a 3D depth sensor, not a laser turret. eufy prints 30,000 Pa across the box, a figure that predicts clean rugs about as well as the last number on a speedometer predicts your commute.

Here is the box against the floor:

The robot is very quiet at work by RTINGS' measure, though the dock is loud when it empties. Upkeep is the chore you'll feel. RTINGS counted more than ten parts that need regular cleaning and called the job very laborious. On hard floors labs found it ordinary: the side exhaust scatters light debris, and corners stay dirty. The mop is its weakest job. Vacuum Wars scored it 82 on dried stains, against that lab's stated average of 108. It's a fair mop for fresh spills and a poor one for last Tuesday's juice.

Will it really steer around dog poop?

Almost certainly, and that is the reason to buy it. Vacuum Wars set out 24 objects and the S2 avoided all of them. RTINGS ran it three times past fake pet waste, and it went around every time. The one blot is The Hook Up, which watched it drive over fake poop in a single pass. The larger samples deserve more weight. The odds are firmly on your side, though no robot is a guarantee. If your dog has an accident while you're at work, this one will very likely leave it for you to find rather than spread it down the hallway.

Is the eufy Omni S2 any good on carpet?

It depends how hard you make it work. At maximum suction over two passes, it pulled 78.8% of a debris mix from high-pile carpet at The Hook Up. That is more than three-quarters of the flour, salt, seed and rice dug out of the deepest pile in the test. Vacuum Wars, also at maximum, got 87% of the sand worked into medium pile, a better haul than most robots manage there. RTINGS ran it as it ships, on Standard suction, and judged its carpet cleaning poor at every pile height. Nobody has rerun RTINGS' test at maximum, so the cause of the gap is unknown. Running flat out also costs battery: RTINGS got over 1.5 hours at maximum and over 3 hours at the lowest setting.

Some trouble doesn't depend on any test. RTINGS saw it get stuck on tasseled rugs, and its mop wets rugs in mixed runs. Then there are the thresholds. When I see a spec sheet quote a climb that two test rigs couldn't repeat, I read it as the best part off a good shift: real once, not something to plan the house around. A step much past an inch and a quarter is a gamble.

That is the S2's one honest catch: it is a hard-floor robot carrying a carpet robot's spec sheet. If most of your home is carpet, our guide to the best robot vacuum for carpet picks the Dreame X60 Max Ultra Complete instead.

Will it keep up with a shedding dog?

On hard floors and short rugs, yes. Loose or flattened hair comes up well, and neither lab's tangle test left any on the brushes. One three-month home test with a dog did see an early hair wrap, which cleared up after firmware updates. Vacuum Wars measured 88.5% of flattened pet hair lifted from carpet, above that lab's average. Hair pressed deep into carpet is another matter, and RTINGS rated it bad there. The side exhaust also blows loose fluff about. The Hook Up watched it puff cotton tufts under the couch. A beagle on wood floors, fine. A husky on wall-to-wall carpet, no.

If shedding is the whole problem and accidents aren't, our pet-hair roundup goes elsewhere. For a carpeted home whose shedders are reliably clean indoors and whose doorways sit under an inch, it picks the Dreame L60 Ultra PE, the robot our own L60 review also recommends for that household. For bare floors with only thin rugs, where every pet is reliably trained, nobody wants a camera aboard and the socks and cables get scooped up before the robot sets off, it picks the eufy Omni C28. The S2 earns its place only where an accident is a real risk.

Were the early S2 problems fixed?

The worrying ones came from hardware you can't buy new. In January 2026 eufy paused preorders for a hardware upgrade it has never described. It sent return labels to buyers who already had robots. Every S2 sold since the late-March US launch is the upgraded version. You may see this called a recall. No regulator issued one, and the CPSC lists no eufy recalls. Vacuum Wars and RTINGS bought theirs at retail after launch.

What isn't known: what eufy changed, and how the robot holds up past its first half-year. The warranty is one year.

Is the firmware still causing trouble?

Less than it was, as far as anyone can tell. Updates arrive every week or two. After firmware 4.1.81 landed in August, one household's robot stopped finishing its route. Version 4.2.3 brought it back to steady. Nothing in those threads mentions pickup changing. No lab said which firmware it tested, so any effect on the carpet split is unproven.

Can it run without a microphone or the cloud?

Mostly, and this is the S2's quiet strength. eufy's user guide lists no microphone, and testers confirm there's no live camera view to hijack. Setup needs the eufy app and an account, and the guide gives no Wi-Fi off switch. After setup, The Hook Up ran it through Matter in Apple Home, room by room, with the internet blocked for 3 days. The weak spot is eufy itself. It publishes no minimum period for security updates. And it admitted in 2022 and 2023 that some of its security cameras weren't encrypted as promised.

So the S2 is not the robot for someone whose first demand is years of guaranteed patches. If your pets are house-trained and long support is the priority, our page on whether robot vacuums are being banned names the Roborock Saros 20, which carries a dated support window eufy doesn't offer. The S2's buyer is different: the worry is a camera that can listen or stream, and the pet occasionally slips up. For that home, it's the camera robot we'd choose.