Buying guides
Home-fit finder
Buying guide
Verdict
Swept's pick under $1000 is the Dreame L60 Ultra PE: the strongest carpet and pet-hair cleaner in this band, for homes whose doorways sit under an inch.
BuyBest for carpet and shedding pets
Opens Amazon.com · Prices change often
At a glance
SuitsCarpeted homes with shedding pets that don't have accidents and thresholds under an inch
The catchStops at a 1 in step; the 1.6 in climb on the box is two steps added together
SuitsWood or LVP homes with flat doorways, rugs that must stay dry and pets that never have accidents indoors
The catchDrives over fake dog poop, and is too tall for low furniture
SuitsHard floors and low rugs with a house-trained shedder, no camera wanted, and someone who picks up first
The catchDodged only 15 of 24 objects, so cords and socks have to go first
SuitsTile or LVP homes without a shedding pet, on a tighter budget, kept tidy
The catchLifted 43% of pet hair pressed into carpet and wraps long hair
Why you see a price band, not a price Prices change daily, so our pages show only the price band a robot sits in and leave the current price to Amazon. That way a page is never wrong by Tuesday.
Two of the labs American buyers lean on hardest name different winners in this band. Neither is wrong. They are weighing different hair. Read every test against every box, though, and one robot does the job most homes here are buying for: carpet and a pet that sheds. Swept's pick under $1000 is the Dreame L60 Ultra PE, for carpeted homes with a shedding pet and doorways under an inch. In the one lab that put it on a scale, not a strand of pet hair stayed in the carpet and nothing wound around its brush. It picked its way past 21 of 24 objects, and Dreame backs it with a 3-year warranty. The catch is the step into the kitchen. It climbs 0.75 in (19 mm) and fails at 1 in (25 mm), whatever the box suggests.
Five things move the pick. If you'd rather answer a few questions, the Home-Fit Finder does it for you.
Every robot picked here is already on US shelves. The FCC's new restrictions don't reach those, so you can buy any of them and keep running it.
BuyBest for carpet and shedding pets
Price band: under $1,000
Dreame's box leads with 30,000 Pa. That figure sells robots and predicts very little. At the brush, this one's suction measured below the lab's average. Yet no pet hair survived in the carpet, and nothing wrapped its brush. When sand was ground deep into medium-pile carpet, it got 94% back out. The pascals do the marketing; the brushes do the cleaning. It dodged 21 of 24 objects, close to flagship avoidance, though no lab has yet put fake poop in its way. A 90-day home test rated its hair pickup and tangle resistance really good. A separate three-week home test never once had to rescue it. The hair figure has a single lab behind it, which is narrower than we'd like. Its carpet, tangle and avoidance results all point the same way, and that's enough for us.
The catch is height. Dreame advertises a 1.6 in (40 mm) climb. The fine print says that's two steps added together, with a gap between them. A single step is rated at 0.9 in (22 mm). The robot cleared 0.75 in (19 mm) and failed at 1 in (25 mm). I spent a career reading supplier specs, and a figure that only works if you use the part twice is a sum, not a spec. If your kitchen sits a full inch above the hall, this robot lives on one side of it. Several long-haired dogs will also test its dock, which empties the bin with a fairly weak pull.
One trap: the L60 Ultra FE (RLL51SE) looks like a cheaper version of this robot. It's laser-only, with no camera, and it dodged 2 of 24 objects on the same course. Make sure the box says PE, model RLL88HE.
The 3-year warranty is three times what anyone else here writes. Dreame's US after-sales policy is less generous in the small print. The battery gets only a year. The cover also applies only to robots bought through Dreame's authorized US channels, so make sure your seller is one.
Buy it if carpet covers most of your floors and your pet sheds, provided no doorway rises a full inch. The full case is in our Dreame L60 Ultra PE review. It's also the carpet pick on Best Robot Vacuum for Pet Hair.
WaitBest avoidance, if dog mess on the floor is the worry
Price band: under $1,600
This is the robot for the morning every puppy owner dreads. Roborock promises its camera and solid-state lidar will see pet waste coming and steer clear. Here the promise holds better than anywhere else in the band. It dodged all 24 objects on Vacuum Wars' course at top sensitivity. RTINGS watched it keep a steady distance from fake pet waste. USB cords are its blind spot. At 3.14 in, it slides under a couch with a 3.1 in gap, and it's very quiet. It also climbs best here, up to 1.5 in (38 mm) at RTINGS.
Pet hair is where the labs part company, and the split has a mechanical cause. The main brush is two short rollers with a gap between them. Hair collects in that gap instead of reaching the bin. A test that weighs what's left on the carpet counts the gathered hair as picked up, and scores it at 92%. Tests that judge one pass, or use real fur, find far less: TechGearLab measured 56%. Unhappy owners agree with the low numbers. With a heavy shedder, the 9 fl oz (270 mL) bin is the weak point; RTINGS saw it overflow and jam before it was full.
We'd wait. Its eyes are as good as any robot we've weighed, and no robot on this page guards better against dog mess. But we won't send a shedding home to it until a lab weighs both the bin and the brush after a hair run. For carpet and a shedder, the L60 Ultra PE dodged 21 of 24 objects and sits in a lower band. The detail is in our Roborock Saros 10R review. If accidents on the floor are your main worry, whatever your budget, our pick for that home is on The Best Robot Vacuum That Avoids Dog Poop, Cords and Toys.
BuyBest if your rugs must stay dry
Price band: under $700
Roborock's first roller mop makes a tidy promise. On carpet, the roller lifts and a shield drops over it, so the rug stays dry. The Hook Up drove seven 2026 flagships across carpet mid-mop, and only this one left the rug dry. It also wiped dried mud off vinyl plank well enough to pass a paper-towel check. For wood floors and one good living-room rug, that matters more than any suction figure.
The mop scores elsewhere look alarming, and the water dial seems to be why. Vacuum Wars gave its dried stains 25, where that review's average is 112, and saw very little water reach the floor. Another reviewer's streaks went away at a higher water setting. It vacuums carpet well, pulling 87% of buried sand, and wraps no long hair. Pressed-in pet hair is weaker, at 77.5%, below that lab's average. At 4.7 in tall, it cleared none of RTINGS' low-furniture gaps, and it climbs only 0.75 in (19 mm).
The catch is the one that would ruin your week. It ran over fake dog poop repeatedly in one test, and RTINGS found it doesn't consistently avoid pet waste. Buy it for wood or LVP floors with flat doorways and rugs that must stay dry, if your pets never have accidents indoors. It's our pick for that reader on Best Robot Vacuum and Mop for Hard Floors.
BuyBest for less, with a shedding pet
Price band: under $500
eufy's page says it "intelligently avoids obstacles". It avoided 15 of 24, which is intelligent in roughly the way a coin toss is. In a three-month home test it ran over fake dog poop and cables, and kept catching on chair legs. That's the bad news, and it's nearly all of it. Elsewhere the C28 cleans like a robot from a higher band. It lifted every flattened pet hair from carpet, pulled 89% of buried sand and wrapped no long hair. Its dried stains scored 119, where that review's average is 108. Its user guide lists no camera and no microphone, and its 12-month warranty matches the Roborocks.
The guide draws the line at rugs deeper than 0.7 in (18 mm), so shag is out. The catch, said plainly: someone picks up cords and socks before every run. Buy it for mostly hard floors and a house-trained shedder, if you'd rather have no camera. It's also the camera-free pick on our page for smaller budgets.
BuyBest for stubborn stains on tile, without a shedder
Price band: under $400
Roborock's page shows an SGS certificate for "0% hair tangling". It's genuine, and it covers the side brush. The main brush wrapped 62% of the long hair in the same lab. Of pet hair pressed into carpet, it recovered 43%, about half what that lab's robots typically manage. It dodged 5 of 24 objects. So far, so poor.
Now put it on tile. On dried stains it scored 149, with 108 as that review's average. On coffee it scored 166, where the average was 94. That's the strongest mopping Vacuum Wars has measured from this Roborock line, from the cheapest band on the page. Its parts list shows no camera or microphone. The manual explains how to switch Wi-Fi off: hold Reset until the voice prompt, then wait five minutes.
The catch is a shedding pet. Hair it misses stays in the rug, and long hair winds around the roller. Buy it for tile or LVP homes without a shedder, on a tighter budget, where the floor gets cleared first.
Roller mops arrived in force this year. The Qrevo Curv 2 Flow, released in January, is Roborock's first, and it has the lab record to be a pick.
The Roomba Max 715 Vac, released in May, is the one to watch, not yet one to buy. iRobot calls it a next-generation upgrade on the Max 705 Vac. The box claims 30,000 Pa, and iRobot says the dock's carbon-lined bags go 90 days between changes. Labs haven't reached it yet. In one reviewer's three-month home test, its dock failed to empty the bin on five tries out of five. Long hair wrapped both its rollers and its new anti-tangle side brush. In fairness, it worked around the clutter course cleanly. We found no owner reports and no written warranty. None of the Max 705's results carry over. If you're a Roomba loyalist who wants a robot with no mop at all, wait for a lab test before buying. For anyone else shopping under $1000, it changes nothing above.
Roborock's newest US launches since July haven't reached the labs yet.
We set each maker's promises against what independent testers and owners found on the exact model, never a sibling or an overseas version. To win, a robot needs proof from two separate directions, such as a lab bench and a long stay in someone's home. The maker's own word never counts as one of them. Then we weighed what matters at home: clutter dodged, pet hair lifted from carpet, long-hair wrap, step height and the written warranty. Nobody measures reliability model by model, so we can't rank it.
The work behind this page covered 8 models, and for 6 of them we had the maker's manual. Our research as a whole keeps 497 sources in full. We also read 148 Reddit threads, which we keep as summaries of what buyers said rather than in full.
They disagree about method far more than about robots. One lab presses short pet hair flat into medium-pile carpet, lets the robot go over it twice, then weighs the hair still on the floor. Another presses synthetic hair into low pile and judges one default pass by eye. The Saros 10R's split brush does well in the first test and badly in the second. The L60 Ultra PE has only been through the first. So compare robots within one lab, then look for agreement across labs.
No. At the brush, the Saros 10R's 22,000 Pa and the L60 Ultra PE's 30,000 Pa measured almost the same, at 0.58 and 0.57 kPa. The difference vanished somewhere between the box and the carpet. Pickup follows brush design and airflow, so ignore the Pa figure.
Measure your tallest single step, not the total of a double one. Only the Saros 10R has been measured clearing more than an inch. The L60 Ultra PE and the Curv 2 Flow stop around 0.75 in (19 mm). The C28 is rated for 0.8 in (20 mm), but nobody has measured it. The Qrevo S Pro has no measured figure. One first-time owner got it over a sill only after marking it as a threshold in the app.
Dreame's 3 years on the L60 Ultra PE, comfortably, with its shorter battery cover in mind. Roborock and eufy write 1 year. We found no current written warranty for either Roomba. Software support is a separate promise. Roborock guarantees security updates for at least 2 years. Its own list ends them for the Saros 10R in June 2027, the Qrevo S Pro in May 2027 and the Curv 2 Flow in February 2028.
| Robot | Best for | Objects dodged, of 24 | What it gives up | Written warranty | Amazon |
|---|---|---|---|---|---|
| Dreame L60 Ultra PE | Carpet and shedding pets | 21 | Fails a 1 in step | 3 years (battery shorter) | View on Amazon (opens in a new tab) |
| Roborock Saros 10R | Avoiding dog mess | 24 | Pet hair on carpet unsettled | 1 year | Wait |
| Roborock Qrevo Curv 2 Flow | Rugs that must stay dry | 16 | Runs over fake poop | 1 year | View on Amazon (opens in a new tab) |
| eufy Omni C28 | Shedder, no camera, for less | 15 | You pick up first | 1 year | View on Amazon (opens in a new tab) |
| Roborock Qrevo S Pro | Stains on tile, no shedder | 5 | Pet hair and long hair | 1 year | View on Amazon (opens in a new tab) |
BuyBest for carpet and shedding pets
It cleans carpet and pet hair as well as anything in this band, without tangling. It looks where it's going, and it carries the only multi-year warranty here. Measure the step into the kitchen first. If it's under an inch, buy it. It keeps the hair off your floors while you're out, and keeps at it after the other warranties have run out.
All lab figures are from Vacuum Wars unless marked. Each of its reviews states its own average, so read every column within this table only.
| Robot | Objects dodged, of 24 | Flattened pet hair, carpet | Long hair wrapped | Buried sand, medium pile | Written warranty | Gives up |
|---|---|---|---|---|---|---|
| Roborock Saros 10R | 24 | 92% (two runs); 56% real fur at TechGearLab | 0% | 80% | 1 year | Pet hair unsettled; small bin clogs |
| Dreame L60 Ultra PE | 21 | 100% | 0% | 94% | 3 years | 19 mm (0.75 in) threshold |
| Roborock Qrevo Curv 2 Flow | 16 | 77.5% | 0% | 87% | 1 year | Dried-stain score 25 (review average 112); cleared no furniture gap at RTINGS |
| eufy Omni C28 | 15 | 100% | 0% | 89% | 1 year | Clutter avoidance |
| Roborock Qrevo S Pro | 5 | 43% | 62% | 81% | 1 year | Pet hair, long hair, clutter |
| Shark PowerDetect AV2820AE | No count published | 89% | 54% | 80% | 1 year | Long hair, battery, app |
| iRobot Roomba Max 705 Vac | Not tested by Vacuum Wars | Not tested by Vacuum Wars | Not tested by Vacuum Wars | Not tested by Vacuum Wars | None found | Dock emptying in question |
| iRobot Roomba Max 715 Vac | Not tested | Not tested | Not tested | Not tested | None found | Untested in any lab |