Does a higher Pa suction rating mean a robot vacuum cleans better?
A robot vacuum's suction rating is the biggest number on its listing, and it keeps climbing. The carpet it has to clean has not changed at all.
On this page4 sections
Roborock’s Saros 20 Sonic carries a 36,000 Pa HyperForce motor among its key features. The older S5 Max, a model that collected awards in 2020, was rated at 2,000 Pa, pitched as enough suction to lift AA batteries. Between the two machines, the biggest number on a Roborock listing grew eighteenfold.
The rest of the category climbed with it. In 2022, 5,000 Pa counted as top tier. By June 2025, Roborock’s Saros 10 and Saros Z70 were rated at 22,000 Pa.
The pascal rating is the figure you are invited to compare first, printed larger than anything else on a listing. If it tracked the dirt coming off your floor, a 2025 flagship would lift more than four times what a 2022 one did. So does it?
What the pascal figure measures
Figure
Where a suction figure is measured
- Sealed portNo brush, no floor, nothing in the way. Pressure reads highest here, at the one point on the airpath where no cleaning happens.
- DuctAfter the bends and the filter. Already well short of the quoted number, with the floor still to come.
- Nozzle, on carpetIn contact with the surface being cleaned. This is the pressure that actually lifts grit, and no specification quotes it.
A pascal is a unit of pressure. On a vacuum it describes how far the fan can pull the air inside the machine below the pressure of the room. Every fan trades that pressure against flow.
Block the inlet and the fan reaches its highest pressure while moving no air at all, like sucking on a straw with your thumb over the end. Open it up and more air moves while the pressure falls. A single pascal figure on a listing doesn’t say where on that curve it was taken.
Cleaning depends on the moving air, because the airstream is what carries loosened grit into the bin. Air watts fold pressure and air volume into one number, and that is how a June 2025 review rates the Dyson 360 Vis Nav. Airflow, its author argued, “comes closer to approximating real-world cleaning power than pascals.”
Leaks count against both. Air drawn in through a worn bin seal or a gap around the brush housing never passes the cleaning head, so the pull at the carpet can sit below what the motor produces.
Whose test produced the number
Roborock footnotes the figure on some of its pages. The S7 is rated at 2,500 Pa, and its footnote reads: “Based on internal testing carried out by the manufacturer according to IEC 62885-2:2016 / 5.8 standards.”
The Qrevo MaxV and its 7,000 Pa point to clause 5.11 of the standard’s 2021 edition, measured in Max+ mode with a full battery. But in September 2026, the 36,000 Pa ratings on the Saros 20 Sonic and the Saros 20 carried no footnote of that kind.
Roborock ran those tests itself, on its own equipment, and no regulation sets how a robot vacuum’s pascals are measured. Yet a test written for robots does exist: IEC/ASTM 62885-7, published in October 2020 as the robot part of the same series on surface cleaning appliances.
Robots in that test drive straight lines over hard floor and carpet, and the result is how much of several kinds of debris they lift, fibers from carpet included. The 2020 test replaced IEC 62929, a robot method from July 2014 that set no safety or performance requirements.
Neither document obliges a maker to run the test or to publish the result. Europe did set pickup floors, for other vacuums: the EU’s 2013 ecodesign regulation for vacuum cleaners required mains-powered machines sold from September 2017 to lift at least three-quarters of the test dust from carpet. Robot vacuums sit on its list of exclusions, beside battery-operated ones.
Seven robots on the same carpet
Picture sand worked deep into a carpet’s pile. It is one of the measured cleaning tests the June 2025 review runs on every robot, built as a stress test of suction, and every robot scores lowest on it. Seven robots’ results sit beside their ratings here.
Hands-on review, June 2025
Listed suction and sand picked up from carpet
| Robot | Listed suction | Sand picked up |
|---|---|---|
| Ecovacs Deebot X2 Omni | 8,000 Pa | 45.6% |
| Dreame X30 Ultra | 8,300 Pa | 31.4% |
| Ecovacs Deebot X8 Pro Omni | 18,000 Pa | 37.7% |
| eufy Omni E28 | 20,000 Pa | 20.6% |
| Roborock Saros 10 | 22,000 Pa | 40.7% |
| Roborock Saros Z70 | 22,000 Pa | 29.1% |
| Dyson 360 Vis Nav | 65 air watts | 53.6% |
The Deebot X2 Omni and its successor, the Deebot X8 Pro Omni, make the plainest pair: the newer robot’s rating is more than twice as high, and it picked up less sand. The two Roborock models share a rating and finished more than 11 points apart.
The eufy Omni E28 carried a higher pascal figure than every robot in the group except the two Roborocks and picked up the least, while the highest result came from the one robot rated in air watts.
As the number of pascals goes up, I’m not necessarily seeing a correlation in real-world cleaning power.
Still, this is one reviewer’s test of seven machines on one kind of debris, and a different floor can reorder them. The Dreame X30 Ultra picked up under a third of the sand, yet its tester wrote that it “excelled in my hardwood tests.” On carpet it could not dig stubborn debris out of the fibers as well as the Deebot X2 Omni.
On carpet, the brush sets the limit
Sand and grit tracked into your carpet settle between the fibers, low in the pile where air passing over the surface barely reaches. Bristles have to knock them loose before the airflow can carry them to the bin. The tester points to the same part: “The shape of the brush roll plays a big part, both in terms of whether it helps agitate the dirt and allows enough airflow.”
Roborock’s own carpet pitch for the Saros 20 Sonic leans on the brush as much as the motor. The pitch describes a brush stabilization module “keeping the main brush in a precisely controlled position, improving carpet-cleaning performance.”
So a higher pascal rating doesn’t tell you which robot will get more sand out of your carpet. The number describes the fan, on a test the maker runs itself. On one sandy carpet, among six robots rated in pascals, the lowest rating picked up the most.
Even the maker of the 18,000 Pa Deebot X8 Pro Omni says as much. Ecovacs’s April 2026 buying guide says the best cleaning results “usually come from a balance between suction pressure and airflow.” A higher Pa rating, it adds, “does not automatically mean better cleaning.”
