0:03 In 2025, robotic vacuums and mops saw a
0:05 ton of new innovations like threshold
0:08 leaping, bagless base stations, roller
0:10 mops, and all new navigation systems.
0:13 But in 2026, we are mostly just seeing
0:15 improvement and refinement of those same
0:18 designs. So, in this video, I've got
0:21 seven flagship vacuums from Roborock,
0:25 Dream, Moova, Yeufy, Narwal, and
0:27 Ecovacs, and we're going to figure out
0:28 if these new models are worth their
0:31 premium price tags or if buying last
0:33 year's model at a steep discount makes
0:36 more sense. And as always, there are no
0:37 sponsored reviews on this channel.
0:39 Starting out with vacuuming performance,
0:42 I mix 10 g each of flour, salt,
0:45 flaxseed, and rice to simulate different
0:47 types of dust and debris, and I sprinkle
0:50 that onto my high-pile carpet. Then, I
0:52 send each vacuum on a two-pass
0:54 vacuum-only run on their highest suction
0:56 power, and I weigh the dustbins both
0:59 before and after the run to determine
1:01 the total amount of debris pickup. And
1:03 of course, I do a thorough vacuuming in
1:05 between runs to pick up any leftover
1:06 debris. And starting out with the least
1:08 expensive vacuum in this year's video,
1:11 the Roborock Q Revo Curve 2 Flow is
1:13 Roborock's first robotic vacuum with a
1:16 roller-style mop, but for vacuuming, it
1:18 includes their Duo Divide brush roller
1:21 system along with 20,000 pascals of
1:24 suction, and the Q Revo Curve 2 Flow
1:26 cleaned for 48 minutes and picked up
1:31 26.9 g or 67.3 percent of the flour,
1:33 salt, flax, and rice mixture while using
1:36 47% of its battery. And after the
1:38 auto-empty process, the Q Revo Curve 2
1:41 Flow had 2.3 g left in its dustbin,
1:43 including a small amount in the bin
1:45 itself, but it was mostly flour caked
1:48 into the air filter. Next, for $1,099,
1:50 is the Moova Mobius 60, which has
1:52 Moova's Trouble Wave double brush
1:55 roller, 30,000 pascals of suction, and a
1:57 a lower gate on the brush housing that's
1:58 supposed to increase the effectiveness
2:00 of its suction power. Unfortunately,
2:02 though, even with all that tech, the
2:05 MoVo only managed to pick up 10.8 g or
2:07 27% of the debris mixture in its
2:10 38-minute run, where it used 42% of its
2:13 battery, and needless to say that is not
2:15 a strong showing. And also concerning
2:17 was after the auto empty process, there
2:20 was still 4.6 g of debris left in the
2:22 bin, which was all heavily caked into
2:23 the air filter, and that significantly
2:25 reduces suction power. After that, for $1,359,
2:27 $1,359,
2:31 is the Dream X60 Max Ultra Complete,
2:33 which has a similar design to the MoVo
2:34 with a suction enhancing gate on the
2:36 brush housing, but it's got a slightly
2:38 redesigned double brush roller and
2:41 35,000 pascals of suction. And the Dream
2:44 X60 Max Ultra Complete cleaned for 28
2:46 minutes, and it picked up 26.2 g or 65.5%
2:48 65.5%
2:51 of the debris, while using just 29% of
2:52 its battery, which is substantially
2:54 better than the MoVo and not quite as
2:56 good as the Q Revo. But concerningly,
2:58 after the auto empty process, the Dream
3:02 X60 still had 7.1 g of dust in its bin,
3:03 and that was a mix of flax and flour
3:05 that was not only caked into the filter,
3:07 but it also had some loose debris in the
3:10 bin. Next, for $1,389,
3:13 is the Roborock S7 MaxV which is the
3:16 2026 version of the Roborock S7 MaxV,
3:18 and while there were some slight changes
3:20 to the Duo Divide brush roller system,
3:22 the biggest upgrade was in suction,
3:25 going from 22,000 pascals on the S7 MaxV
3:28 to 36,000 pascals on the S7 MaxV. But in
3:30 my years of testing vacuums, I've found
3:32 that carpet pickup is mostly about brush
3:35 roller design rather than suction. And
3:37 the S7 MaxV cleaned for 24 minutes,
3:39 which was the fastest by far, and it
3:41 only used 21% of its battery, while
3:46 cleaning 25.1 g or 62.8% of the flour,
3:48 salt, flax, and rice mixture, which is
3:51 actually 3 g less than last year's S7
3:53 MaxV despite the increase in suction.
3:55 And that puts the Saros 20 in third
3:56 place so far. And after the auto empty
3:59 process, the Saros 20 had 1.9 g left in
4:02 its bin, which again was mostly flour
4:03 caked into the air filter. After that,
4:06 for price of $14.99, is the just
4:08 released Narwal Flow 2, which, as the
4:10 name implies, is the successor to last
4:12 year's Narwal Flow, which was the
4:14 highest performing mop that I've ever
4:16 tested. But it did struggle with carpet
4:18 vacuuming performance. And this year,
4:20 the Narwal Flow 2 cleaned for 25
4:23 minutes, and it used 24% of its battery
4:26 to pick up 18.6 g of debris, which is 46.5%
4:27 46.5%
4:30 of the 40 g possible. And while that's
4:32 6% higher than last year's Narwal Flow,
4:34 it still puts the Flow 2 in second to
4:36 last place so far. However, after the
4:38 auto empty process, the Narwal Flow 2
4:41 had just 0.1 g left in its dust bin,
4:44 which is extremely impressive, and the
4:46 air filter was almost completely clear.
4:47 And that is important because it's going
4:49 to help prevent the Narwal Flow 2 from
4:51 losing suction over time. After that,
4:53 we've got the $1,499
4:57 Ecovacs X12 Omni Cyclone. And the X12's
4:59 single brush roller remains mostly
5:01 unchanged from last year's X11. But the
5:04 X12 did get a slight boost in suction
5:08 from 19,500 pascals up to 22,000. And in
5:11 this year's test, the X12 Omni Cyclone
5:14 cleaned for 46 minutes, and it used 59%
5:17 of its battery to pick up 17.9 g, or
5:20 44.8% of the debris mixture, which puts
5:22 it just slightly behind the Narwal Flow
5:24 2. But it is worth noting that it
5:26 cleaned for almost twice as long, and it
5:28 used over twice as much battery to do
5:29 that. However, even though it had lower
5:31 than average pickup, it had much higher
5:33 than average auto empty efficiency,
5:36 leaving just 0.8 g in its dust bin after
5:38 the auto empty process. And it did have
5:40 some flour caked in the filter, but much
5:42 less than the rest of the vacuums, other
5:43 than the Narwal. And last, the most
5:45 expensive vacuum in this video is the
5:48 brand new Yeufy Omni S2, which shares
5:51 the same general shape as the U S1 Pro,
5:54 which was my top pick in 2024, but the U
5:56 S2 is actually quite a bit bigger and it
5:58 has a completely redesigned divided
6:00 brush roller, a dust bin that resembles
6:02 a stick vacuum, and 30,000 pascals of
6:05 suction. And in my carpet pickup test,
6:08 the U S2 cleaned for 34 minutes, it used
6:10 52% of its battery, and it picked up a
6:12 total of 31.5
6:14 or 78.8%
6:16 of the debris mixture, which not only
6:18 catapults it into first place in this
6:20 video, but it's also the highest score
6:22 that I've recorded since switching to my
6:24 current testing equipment 3 years ago.
6:25 And after the auto empty process, the U
6:28 did have 1.5 g of debris left in its
6:29 bin, but its main air filter was
6:31 completely clean thanks to its unique
6:33 dust bin design. And if we compare this
6:35 year's results to all the other vacuums
6:38 that I've tested, the U S2 sits on top,
6:39 but the rest of the vacuums in this
6:41 video fall significantly below the more
6:44 budget-friendly Moovo P50 and Dream L40
6:47 S Ultra C E that previously held the top
6:48 two spots for carpet vacuuming
6:50 performance. Next, moving on to hard
6:52 floor vacuuming, I used that same 40 g
6:55 mixture of flour, salt, flax, and rice,
6:57 and historically vacuums score really
6:59 well in this test between 95 and 100%
7:01 pickup. But something that I've never
7:03 really reported on before is the feel of
7:05 the floors on my bare feet after
7:07 cleaning because sometimes even vacuums
7:09 with a high pickup percentage leave the
7:11 floors feeling gritty. So, in addition
7:12 to measuring their actual pickup
7:14 performance, which you can see here with
7:16 all the vacuums scoring between 96 and 99.5%,
7:18 99.5%,
7:20 I also gave each vacuum a more
7:21 qualitative score for how the floors
7:23 felt after the cleaning. And some
7:25 notable results were the Roborock Q Revo
7:27 Curve 2 Flow that had a
7:29 middle-of-the-road pickup score due to
7:30 the fact that it knocked some rice
7:32 grains under my refrigerator, but it
7:34 still had a very clean floor feeling.
7:36 And on the other end of the spectrum,
7:39 the Moovo Mobius 60 picked up 99.3%
7:41 of the debris mixture, but it left a
7:43 thin film of flour dust on the floor,
7:45 which made the floors feel pretty dirty.
7:46 And if we combine the carpet and the
7:48 hard floor scores together, you can see
7:50 that the Ufi S2 finishes in first with
7:53 an 88.6% combined score, which does put
7:55 it barely in the first place for all the
7:57 vacuums that I've ever tested, but it
7:59 was definitely held back by its slightly
8:00 lower than average hard floor pickup
8:02 score, which was caused by its side
8:04 facing exhaust that can push debris out
8:06 of the way as it passes from left to
8:08 right. However, just looking at their
8:09 total pickup performance doesn't tell
8:11 the whole story. Because one of the big
8:13 reasons why less expensive vacuums have
8:15 better carpet pickup performance is that
8:17 unlike flagships, they aren't as focused
8:19 on avoiding hair tangles, and therefore
8:21 they have completely different brush
8:23 roller designs. So, next to test their
8:25 hair pickup and tangle avoidance, in my
8:27 initial test I counted just 30 strands
8:29 of 12-in human hair and I spread them on
8:31 my bathroom floor before sending each
8:34 vacuum on a single pass vacuum only run.
8:36 And then I carefully inspect both the
8:37 floor and the vacuum to find any
8:39 leftover or tangled hair. And in this
8:41 test, all the robots completely cleaned
8:43 the hair off the hard flooring, and only
8:46 the Mova M60 and Dreametech S60 Max
8:48 Ultra Complete had a single hair on the
8:51 vacuums themselves. But my second test
8:53 is a lot more of a stress test, where I
8:56 spread 1 g of 16-in synthetic blue hair
8:58 and two cotton balls separated into
9:01 tufts to simulate pet fluff on my high
9:03 pile carpet. And after the vacuuming
9:04 run, I collect any hair that's left on
9:06 the carpet and I inspect the robots
9:08 themselves for tangles. And the best
9:11 performer by far here was the Narwal
9:14 Flow 2, which cleaned 100% of the hair
9:16 and left just a single tuft of fluff on
9:18 the carpet. And it also managed to stay
9:20 completely tangle-free with just one
9:22 little tuft of cotton stuck on its
9:23 wheel. And I originally thought that the
9:26 Ufi Omni S2 also got a perfect score
9:28 with nothing left over on the carpet or
9:30 on the robot, but after reviewing the
9:31 footage, you can see that the side
9:34 facing exhaust acted like a leaf blower
9:35 and it actually just pushed all the
9:37 fluff under the couch instead of picking
9:39 it up. But, it did do a good job picking
9:41 up the blue hair, and it also ended up
9:42 being completely tangle-free on the
9:44 vacuum itself, which is still a pretty
9:46 good result. The Moovo Mopious also
9:48 performed decently in terms of hair
9:50 pickup, leaving just a few broken blue
9:52 hairs and a small bundle of fluff on the
9:54 carpet, but it had a large tangled ball
9:56 of hair stuck in its wheel well. The
9:59 Roborock S7 MaxV ,Qurivo Curve 2 Flow,
10:01 and Dream X60 Max Ultra Complete all had
10:03 pretty average results in terms of hair
10:05 pickup and tangling, but the Ecovacs X12
10:07 Omni Cyclone really struggled, and it
10:09 not only left a significant amount of
10:11 hair and fluff on the carpet, but it
10:13 also failed to get any of that hair into
10:15 its dustbin, and it left a matted mess
10:17 clogged behind the brush roller. I also
10:19 tested the vacuum's corner vacuuming
10:21 ability by placing 1 tsp of dried bread
10:23 crumbs in four different hard-to-reach
10:25 locations, and I awarded them 1 point
10:28 for a partial clean and 2 points for a
10:30 completely cleaned pile. And in this
10:33 test, the eufy Omni S2 and Roborock S7
10:35 MaxV scored six out of eight possible
10:37 points by completely cleaning three of
10:39 the four piles, but they failed to clean
10:41 under the refrigerator. While the Dream
10:43 X60 Max Ultra Complete also scored six
10:46 points by completely cleaning two piles
10:47 and partially cleaning both the inside
10:49 cabinet corner and under the
10:51 refrigerator. The Qurivo Curve 2 was
10:53 held back here because it was too tall
10:54 to fit under the counter stool,
10:56 resulting in a score of four out of
10:58 eight points. The Ecovacs X12 completely
11:00 skipped the pile by the dishwasher, and
11:02 it scattered the pile on the inside
11:03 corner for a score of three out of
11:05 eight, and the Moovo decided that it
11:06 couldn't fit under the stool, even
11:08 though it has a retractable lidar
11:10 turret, and it completely missed the
11:12 inside corner pile for a score of two
11:14 out of a possible eight points. And if
11:16 we then combine all the vacuuming scores
11:19 together, that puts the eufy Omni S2 on
11:21 top, and it's basically only held back
11:23 by that singular issue that its powerful
11:25 suction motor exhausts out of the side
11:26 of the vacuum instead of the back, and
11:28 that gives it a leaf blower effect on
11:30 extra light debris. But, the Omni S2
11:32 recorded the highest ever carpet
11:34 vacuuming score. It had excellent hair
11:37 tangle avoidance and a great corner
11:38 cleaning algorithm. And the surprising
11:41 second place vacuum was the lower priced
11:43 Roborock Q Revo Curve 2 Flow, which had
11:45 above average carpet and hard floor
11:48 pickup, good tangle avoidance, and above
11:50 average corner cleaning ability, but it
11:51 was too tall to fit under my stools,
11:53 which knocked down its overall corner
11:55 pickup score. And then third place was a
11:58 three-way tie between the Dream X60 Max
12:00 Ultra Complete, the Narwal Flo 2, and
12:03 the Roborock Saro S20, which were mostly
12:05 just average when it comes to vacuuming
12:07 performance. So next we need to look at
12:08 their mopping performance. And in my
12:10 first test, I spread out 10 ml of
12:13 Florida mud on my LVP flooring, and I
12:15 dry it with a hair dryer before sending
12:17 the robots out on a two-pass combined
12:19 vacuum and mopping run with the mopping
12:21 water level set to maximum. And after
12:23 the run, I spray the stain with water,
12:25 and I wipe it clean with a paper towel,
12:27 and then I set the towels aside to dry.
12:28 And in this video, there are two
12:31 different mopping types. The Roborock
12:34 Saro S20, Moova Mop 60, and Dream X60
12:37 Max Ultra Complete all use dual spinning
12:39 mopping pads that periodically clean
12:41 themselves at the base station. While
12:44 the Ecovacs X12 Omni Cyclone, Narwal Flo
12:47 2, Roborock Q Revo Curve 2 Flow, and Ufi
12:50 Omni S2 all use roller style mops that
12:52 are constantly being cleaned by wetting
12:54 the roller with clean water and then
12:56 scraping the dirty water out into an
12:58 onboard dirty water storage tank. The
13:00 Moova Mop 60 also has another trick
13:02 where it actually has three different
13:04 sets of pads that it can interchange for
13:06 different flooring types, but in my
13:07 testing, it seemed like there was only a
13:09 very minor difference in the cleaning
13:11 ability between the three pad types. And
13:13 I would have loved to see a more
13:15 abrasive mopping pad option included for
13:17 more durable flooring types. The Ecovacs
13:20 X12 Omni Cyclone also has a very unique
13:22 feature where it can spot stains on the
13:23 floor and spray them with its
13:25 front-facing water jets to soak them
13:28 before mopping. But in my mopping test,
13:30 the X12 did not identify the large mud
13:32 stain as an area that needed to be
13:34 pre-wet with its jet spray.
13:36 And looking at the final clean stains,
13:38 the Dream X60 Max Ultra Complete had the
13:40 hardest time and really only cleaned the
13:43 stain on the seams of my LVP. The Moova
13:45 Mop 60 was next and it also struggled to
13:47 clean the middle of each plank. The
13:49 Ecovacs X12 Omni Cyclone performed
13:51 surprisingly poorly for a roller mop.
13:53 The Roborock S20 was slightly better
13:56 than that. The Yeufy Omni S2 did an
13:58 overall decent job, but still left
14:00 noticeable mud on the floor. The Narwal
14:02 T2 Flow did very well and the floor was
14:04 visibly clean, but the Roborock Q Revo
14:07 Curve T2 Flow did by far the best job
14:09 and it left a floor that was not only
14:11 visibly clean, but it also passed with a
14:13 completely clean paper towel test. And
14:15 speaking of that paper towel test,
14:16 here's what those look like for this
14:19 round of mops. And out of all the robots
14:20 in this video, I would say that I'm most
14:22 disappointed in the Ecovacs X12 Omni
14:25 Cyclone that didn't activate its jet
14:27 spray system at all during this run. And
14:29 just out of curiosity, I re-ran the test
14:32 in its forced jet spray mode and the
14:33 results were excellent, finishing
14:35 somewhere in between the Narwal Flow T2
14:38 and the Q Revo Curve T2 Flow for both
14:40 visual score and paper towel test. But,
14:41 Ecovacs really needs to update their
14:43 firmware to get the jet spray more
14:45 active in normal mopping modes. I also
14:47 tested the corner mopping ability of
14:48 these robots by putting a smear of hot
14:50 sauce in five separate hard-to-reach
14:52 locations in my kitchen. And like the
14:55 corner vacuuming test, I give each robot
14:57 1 point for a partial clean and 2 points
14:59 for fully cleaning each spot. But,
15:01 importantly, I award no points for only
15:03 cleaning the stain with the side brush,
15:04 only for cleaning it with the mops. And
15:06 the very surprising worst performer on
15:08 this test, scoring just 2 out of 10
15:11 points, was the Yeufy Omni S2, which
15:13 completely missed every single corner
15:15 stain and it bumped into and moved the
15:17 stool when it was cleaning under it,
15:19 which is really surprising since the
15:21 Yeufy S1 Pro was one of the most
15:23 confident navigators that I have ever
15:25 tested. And it seems like the S2 doesn't
15:27 understand its own dimensions, which
15:29 makes it clumsy and affects its ability
15:31 to properly hug the corners with its
15:34 extendable mop. The Narwal Flow 2 also
15:36 had issues with clumsy navigation, and
15:38 despite having a new extendable mop, it
15:39 still struggled to adequately clean
15:42 corners, scoring five out of 10 possible
15:45 points. The Ecovacs X12 Omni Cyclone ran
15:47 its side brush through every single
15:49 stain, but it only partially cleaned
15:50 most of those stains with its mop,
15:52 scoring six out of 10 points. The
15:54 Roborock S7 T20 could have scored
15:55 significantly better if it used its
15:57 extending mopping pad a little bit more
15:59 liberally, but it ended up only scoring
16:02 six out of 10 points. The Mova Mop 60
16:03 also had a bunch of partial clean
16:06 scores, earning six out of 10 total. But
16:07 on a positive note, it did decide to
16:09 clean under the counter stools this
16:11 time, unlike in the corner vacuuming
16:12 test when it determined that it couldn't
16:14 fit under them. And the top score went
16:16 to the Dreame X60 Max Ultra Complete
16:18 that did use its extending mopping pad
16:20 liberally, and it was slow and
16:22 methodical around corners, fully
16:24 cleaning two of the stains and partially
16:26 cleaning the other three for a score of
16:27 seven out of 10. And the last thing that
16:29 I test with mopping is whether the
16:31 robots can perform a combined vacuuming
16:33 and mopping run without getting the
16:35 carpets wet in a room with mixed
16:36 flooring. And out of all the robotic
16:38 mops in this video, only the Roborock
16:40 Qurevo Curve 2 Flow was able to
16:42 completely avoid getting the carpets wet
16:44 by using a retractable plastic cover
16:46 that goes over the roller mop. None of
16:48 the robots that had spinning mopping
16:49 pads were able to lift them high enough
16:51 to avoid dragging them on the carpet,
16:53 but all three of them are able to drop
16:55 their mop pads at the base station and
16:57 prioritize carpet cleaning first. So,
16:59 the only time that mop lifting should
17:01 actually matter is if the vacuum has to
17:03 pass over a carpeted area to get from
17:05 one hard floor to another. Neither the
17:07 Narwal Flow 2 nor the Ufi S2 could lift
17:09 their roller mops high enough to avoid
17:11 getting the carpets wet, and neither of
17:13 them has a plastic cover for their mop
17:15 roller like the Qurevo Curve 2 Flow. And
17:17 while the Ecovacs X12 Omni Cyclone does
17:18 have a retractable cover for its roller
17:20 mop, it was still getting the carpets
17:22 really wet, and I was able to figure out
17:23 the cause, which is that the back of the
17:25 robot gets soaked during mopping, and it
17:27 then transfers all that water onto the
17:29 carpet. So, then if we average the three
17:31 mopping categories, you can see that the
17:33 Roborock Q Revo Curve 2 Flow comes out
17:35 on top with by far the best actual
17:37 mopping score, the best carpet crossing
17:39 score, and a middle-of-the-road corner
17:41 mopping score. And while the Narwal Flow
17:43 2 and Ufi S2 did a relatively good job
17:45 with stain cleaning, their poor corner
17:47 cleaning performance and inability to
17:49 cross carpets negatively affected their
17:51 overall scores. Next, moving on to
17:53 object avoidance, I set up a room with a
17:55 pair of untied shoes, a loose extension
17:58 cord, and a pile of plastic dog poop,
17:59 and I awarded the robots 1 point for
18:01 touching either the shoes or the
18:03 extension cord without getting stuck, 2
18:05 points for avoiding them completely, and
18:08 an all-or-nothing 2 points for avoiding
18:09 the dog poop. And the top performer
18:11 here, with a perfect six out of six
18:14 score, was the Ecovacs X12 Omni Cyclone,
18:16 which is a major victory for Ecovacs,
18:17 which has traditionally struggled with
18:20 this test. The Moovo Miboia 60 almost
18:22 scored a perfect score, avoiding every
18:24 object for nearly 45 minutes before
18:26 accidentally getting too close to the
18:27 dog poop at the end of its run, ending
18:29 with a total of four out of six. The
18:32 Dream X60 Max Ultra Complete ran
18:33 directly over the dog poop at the
18:34 beginning of the run, and then
18:36 successfully avoided it for the
18:38 remainder of its cleaning task for a
18:40 score of four out of six. And the
18:44 Roborock S7 20, Ufi S2, and Narwal Flow
18:46 2 all ran directly over the dog poop and
18:48 gently nudged the shoes for a score of
18:51 three out of six. While the Roborock Q
18:53 Revo Curve 2 Flow showed why it is
18:54 considered less of a flagship level
18:57 vacuum, and it had a lot of trouble in
18:58 this test, bumping into the shoes
19:00 multiple times, completely ignoring the
19:02 loose cord, and running over the dog
19:05 poop over and over again for a score of
19:07 two out of six points. Another thing
19:08 that we're seeing a lot more often on
19:10 flagship vacuums is their ability to
19:13 cross very tall thresholds. So, I tested
19:15 three different heights of threshold,
19:17 each with a ramp side and a flat side.
19:20 And in this test, the Roborock Saro 20,
19:22 Moovo Mobius 60, and Dream X60 Max Ultra
19:24 Complete were all able to traverse the
19:27 giant 50 mm threshold in both
19:30 directions. The Narwal 2 Flow and Ufi S2
19:31 were able to traverse the 30 mm
19:34 threshold in both directions, and both
19:36 the Ecovacs X12 Omni Cyclone and
19:38 Roborock Q Revo Curve 2 Flow were able
19:40 to cross the ramp side of the 30 mm
19:42 threshold, but they failed to cross the
19:45 flat side. And all the vacuums easily
19:47 crossed the more reasonably sized 13 mm
19:49 threshold. And the last thing to look at
19:50 on the robots themselves before we move
19:52 on to their apps and smart home
19:54 platforms is their noise level in case
19:55 you want to have these things running
19:57 while you're in the house. And in this
19:58 test, I set each vacuum on their second
20:00 highest suction power, and here they are
20:31 And the final thing that we need to
20:33 cover are the apps and smart home
20:35 integrations for these robots. And for a
20:37 long time, the apps were mostly all the
20:39 same. But as features have gradually
20:42 been added, the Roborock app has risen
20:44 to the top, being both user-friendly and
20:46 feature-rich. Recent changes to the
20:48 Moovo and Dream apps have made the maps
20:50 less detailed in favor of a more
20:52 square-based design that makes it harder
20:54 to set up no-go zones and recognize
20:55 features in your home. And as they
20:57 continue to add more and more features
20:59 and options, they have become buried in
21:01 submenus that makes them more difficult
21:04 to understand, find, and use. Similarly,
21:06 the Ecovacs app has a lot of options
21:08 that get buried deep in the menu and
21:10 it's put a lot of emphasis on their AI
21:12 agent mode and their Yiko voice
21:14 assistant, which is fine, but if I'm
21:16 accessing the app, it's probably because
21:18 I want to change a setting, not talk to
21:20 AI. The Narwal app is perfectly fine and
21:21 it actually has a pretty manageable
21:23 number of options, but I don't like that
21:25 the default option for each of the
21:27 cleaning tasks is their automatic Freo
21:29 mode since in my experience those modes
21:31 can take an extremely long time to
21:32 complete if the robot decides that it
21:35 needs to vacuum or mop a room several
21:37 times. But by far my number one gripe
21:39 with the Narwal app is that by default
21:42 carpets are set to cross only, which
21:43 will prevent the Narwal Flow 2 from
21:45 cleaning your carpets at all even when
21:47 it's in vacuum only mode. And last, the
21:49 Ufi app is as simple as it could
21:52 possibly be and the 3D house map works
21:54 well and it's easy to edit. And setting
21:56 up tasks, like I said, is extremely
21:58 simple. But it is the one app that
22:00 doesn't have enough options. And as far
22:02 as I can tell, it doesn't even have an
22:04 option to vacuum carpets first before
22:06 getting the mop wet, which is something
22:07 that most other brands have been able to
22:09 do for years. And out of the seven
22:11 vacuums in this video, Ufi is also the
22:13 only one that doesn't allow you to use
22:15 the front-facing camera on the robot to
22:17 drive it around the house remotely,
22:18 which again is a feature that most
22:20 flagship options have had for years. Ufi
22:21 is also the only one that doesn't
22:23 support room-specific commands when
22:25 you're asking it to clean via Amazon
22:27 Echo and it doesn't have any built-in
22:29 voice assistant, which I actually don't
22:30 mind, but it seems like something that
22:32 the most expensive vacuum in the video
22:34 might have. However, none of this
22:36 matters because what matters is matter
22:38 and it's here and for the first time
22:39 almost all these vacuums are
22:42 controllable locally when connected via
22:44 matter over Wi-Fi. And for my testing, I
22:46 connected the vacuums to Home Assistant
22:48 via matter, but I was disappointed to
22:50 find that there wasn't room-specific
22:51 controls. And it turned out that it
22:53 actually was possible, but it involved
22:55 first mapping the vacuum segments to
22:57 Home Assistant areas and then calling
23:00 the clean area with vacuum cleaner
23:01 action. And all that seemed a little bit
23:03 cumbersome, so I went searching for a
23:04 better way. And after I shared the
23:06 vacuums with Apple HomeKit, I could
23:09 select vacuuming, mopping, or vacuum and
23:11 mop, and I could select any number of
23:13 the rooms that I had on the map. I also
23:15 used my router to block all the vacuums
23:16 from the internet, and amazingly, they
23:18 remained completely controllable via
23:20 HomeKit. And whether they can remain
23:22 disconnected indefinitely is still up in
23:24 the air, but as of recording this video,
23:25 they've been blocked for 3 days, and
23:27 they're all still going strong. And that
23:29 is really exciting because this is the
23:31 first time that we've had pure local
23:33 control with high usability on flagship
23:35 robotic vacuums. So, that means that now
23:37 it is conclusion time, and in this
23:38 chart, I have averaged all the
23:41 individual grades for each category with
23:43 carpet vacuuming and hard floor mopping
23:44 given double weight. And you can see
23:46 that the Roborock Quriovo Curve 2 Flow
23:48 that had the second highest carpet
23:50 vacuuming score and the best mopping
23:52 performance ended up in first place. And
23:53 don't forget that the roller mop cover
23:56 was 100% effective at keeping carpets
23:58 dry, making it an excellent option for
24:00 homes that have mixed hard flooring and
24:02 high-pile carpets. And it is a great
24:04 pick and an excellent value considering
24:05 its price much lower than the rest of
24:07 the vacuums in this video. However, you
24:09 should be aware that it is missing some
24:11 common flagship features like advanced
24:13 object avoidance, high threshold
24:15 crossing, and it is much taller than the
24:19 Roborock S7 20, eufy S2, and Dream X60
24:21 Max Ultra Complete, so it's not as good
24:23 at cleaning under low furniture. And to
24:24 me, while making this video, the eufy
24:26 Omni S2 seemed like the most
24:28 flagship-level vacuum, recording the
24:30 highest carpet vacuuming score of any
24:32 vacuum that I've ever tested with well
24:34 above-average mopping performance, great
24:36 hair tangle avoidance, good object
24:39 avoidance, and Matter connectivity. And
24:40 its two biggest shortcomings are the
24:42 fact that it has a side exit exhaust
24:44 that pushes light debris out of the way,
24:46 and it's got that corner mopping
24:48 algorithm that makes the Omni S2 look
24:49 like it has no idea what its own
24:51 dimensions are. But both of those issues
24:53 could be easily fixed with better path
24:54 planning that can happen via firmware
24:56 updates. And while they're at it, eufy
24:58 should also add an option to vacuum all
25:00 the carpets first before getting the mop
25:02 wet, since it also lacks a roller mop
25:04 cover like the less expensive Roborock Q
25:07 Revo Curve 2 Flow has. If you definitely
25:09 need a threshold climber, both the
25:12 Roborock S8 20 and Dreametech X60 Max
25:14 Ultra Complete easily cleared even the
25:16 tallest 50 mm thresholds. But while
25:17 neither vacuum had any major
25:19 shortcomings, they mostly just performed
25:21 average or slightly above average in all
25:23 of my tests, which is pretty
25:24 disappointing for newly released
25:27 flagships with very premium price tags.
25:28 And if you do have tall thresholds in
25:30 your house, it might make sense to look
25:32 at last year's Dreametech X50 Ultra that
25:35 can also clear 50 mm thresholds and had
25:37 similar vacuum and mopping performance
25:40 for $500 less. As always, there are no
25:41 sponsored reviews on this channel, but I
25:43 do have links to all the vacuums in this
25:45 video down in the description. And as
25:46 always, I appreciate it when you use
25:47 those links, since as an Amazon
25:49 affiliate, I do earn a small commission
25:51 on the sale at no cost to you. I'd also
25:53 like to thank all of my awesome patrons
25:55 over at Patreon for their continued
25:56 support of the channel. And if you're
25:57 interested in supporting this channel,
25:58 please check out the links down in the
26:00 description. If you enjoyed this video,
26:02 please consider subscribing, and as
