SwitchBot S20: two-minute review

The SwitchBot S20 is a hybrid robot vacuum with a particularly impressive mopping setup. While most robot vacuums feature flat mop pads that tend to spread spillages around, the S20 is equipped with a roller that's fed a continuous stream of fresh water, while any dirt is scraped off as it goes. The model also arrives with a comprehensive dock that not only empties the bot's onboard bin, but drains its dirty water tanks, refills the clean water tank and dries the mop pads, too.

I tested it out and was generally impressed, although a few quirks prevent me from recommending it as one of the best robot vacuums on the market right now. The mopping works as advertised, providing a more thorough clean than I've seen with most bots. Vacuuming is solid, too, while object detection and avoidance are perhaps the best I've encountered.

However, the S20 gets into a pickle if set to vacuum after it has completed a mopping run, when its roller is still a bit damp. In this situation, any debris on the floor is spread around, and no discernible vacuuming takes place. That's okay if your home's layout makes it easy to keep vacuuming and mopping completely separate (with the mopping last on the list), but I wouldn't recommend the S20 as a bot that can transition smoothly between the two tasks.

It's also rather noisy. I'd say it's louder than average in every cleaning mode, but the sound of it auto-emptying is especially alarming. For navigation the bot performed well, but did prove occasionally erratic – perhaps down to that ultra-sensitive object avoidance – and the app could be improved with a few tweaks.

At full price, this bot sits in the premium bracket, but hefty discounts are common and deliver much better value for money. If mopping is your priority and you spot a decent deal, the S20 could be a great buy.

A final note before I begin with my full SwitchBot S20 review. If you're here because you heard that SwitchBot makes mini robot vacuums, be advised that the S20 isn't one of those bots. The robot itself and the dock are full-sized, and also a little chunkier than average. Check out the SwitchBot K11+ if you're after a tiny home-helper that can tuck into tight spaces.

SwitchBot S20 review: price & availability

• List price: $799.99 / £799.99
• Launched: September 2025
• Available: US / UK

The SwitchBot K20 launched in September 2025, with a list price of $799.99 in the US and £799.99 in the UK. At time of writing, there was a voucher on the SwitchBot US site to activate a 44% discount (taking it to $448) and one for 39% off on the UK sites (taking it down to £488). Given the availability of big discounts so close to launch, it seems likely that this bot will not be sold at full price very often.

There's also an "Auto-Fill & Drain System" version of this model, which you plumb into your home's water supply. It costs the same at ticket price.

At full price this is a premium robot vacuum, but those chunky discounts see it sit more middle of the pack. On paper, at the discounted price it's a decent value proposition, but just make sure you're happy to live with the various quirks before you take the plunge.

• Value for money score: 3.5 out of 5

SwitchBot S20 specs

SwitchBot S20 review: design

• Rather large, hybrid robot vacuum and mop
• Roller mop that's continually cleaned as it mops
• Dock empties dust, dries roller and drains/refills the onboard tanks

SwitchBot is known for making tiny little robot vacuums, such as the dinky SwitchBot K11+, so I was quite surprised when I unpacked the K20 to find that it wasn't small at all. In fact, far from being scaled-down, this bot is on the chunkier side for a modern robot vacuum.

By way of example, the Roborock Qrevo Master (a not especially swish or streamlined robovac) measures 13.9in / 35.3 cm in diameter and 4.1in / 10.3 cm tall, whereas the K20 is 14.4in / 36.5cm wide and 4.5in / 11.5cm tall.

The headline feature is the mop setup. The K20 has a roller mop that applies 1kg of pressure to scrub dirt from your hard floors. The roller is continually fed with clean water as it rolls across your floor, while a scraper siphons off the grime and dirty water and holds it in a small secondary water tank. While the roller can't kick out to the side of the machine for edge cleaning, as some premium models offer, it's still a very modern mop setup.

The suction stats are good, too, topping out at 10,000Pa. Until recently, that was about the maximum you'd find anywhere. Nowadays, there are premium bots that can manage twice that; but 10,000Pa should still be ample for most people's needs. There's a fixed side-sweeper brush and a single rubber roller.

Many modern robovac brands are exploring ways to help their bots scale tall room thresholds, but not so much SwitchBot. The chunky wheels should enable the S20 to make its way over steps up to 0.8in / 2cm – and perhaps a bit over that, based on my tests – but beware if your home has thresholds beyond that figure.

On the top side of the S20 you'll find three physical buttons. One is your on/off, another sends the bot back to its dock, and the third I'm not quite sure about. In any case, you'll generally be controlling your bot using the companion app. It's also Matter-compatible, so you can control it using Alexa, Google Assistant and so on.

There's also the LiDAR turret – or rather the LiDAR SLAM, as a label informs me. SLAM is an older, less-efficient method of robovac navigation (here's more on LiDAR vs SLAM in robot navigation), but there's no mention of it anywhere else in the literature. This is bolstered by an AI camera mounted at the front, for object identification and avoidance.

The dock is designed to be able to take care of a number of maintenance tasks for you. It sucks the dust from the small onboard bin into a larger dust bag, drains the bot's dirty water tank and refills its clean one, and washes the roller and dries it with warm air (122ºF / 50ºC).

Unlike other similar models I've tested, the bot here docks only a short way into the base station, rather than driving right inside. I'm not sure exactly how it dries because the roller is hidden under the bot, away from the dock.

• Design score: 4 out of 5

SwitchBot S20 review: performance

• Excellent mopping and solid vacuuming, but a little noisy
• Sensitive object avoidance; occasionally erratic navigation
• Vacuuming with damp mop disastrous

Setup for the S20 was fairly standard. I performed a "quick mapping" run, where the bot ventured into all the rooms it could find, looked around a bit (without trying to clean them) and started to put together a map. No major issues here, although the bot did miss a room (I was confident it would find it later, and it did). I then edited the map and we were ready to start cleaning.

In terms of cleaning options, you can vacuum and mop together, just vacuum, or vacuum then mop. There are four suction modes (Quiet, Standard, Strong and Max) and Moist or Wet mopping modes.

Vacuum performance

To kick things off, I sent the bot off on a whole-floor vacuum, using the Strong suction mode, to get a general feel of its performance. My initial reaction was that it seemed noisier than other bots I'd used, with a vacuuming pitch that was a little irritating.

On the plus side, though, I could hear it ramping suction up and down as it moved from hard floor to carpet. When I paused and checked the onboard bin, I found plenty of dust and dog hair in there. The following auto-empty process was the loudest I've experienced in all my time reviewing robot vacuums.

Now on to TechRadar's standard suction tests. I sprinkled a mix of dry oats, lentils (to represent chunky debris) and dry tea leaves (to represent fine debris) on a carpeted floor, and asked the bot to clean the area twice, in Strong suction mode.

It performed fairly well, sucking up the lentil/oat mix efficiently, and clearing a good amount of the tea as well. I followed up with an area clean in Max mode, which removed even more tea, leaving only a light sprinkling. This is a solid result – it's a purposefully tricky test.

On hard floor, the test didn't go quite as smoothly. I ran my first hard floor test after I'd used the S20 for some mopping, and it was a disaster – by the time the S20 made its proud "Cleaning complete" announcement, there was tea, oats and lentils all over my kitchen floor.

Repeating the test on a different day, with the mop roller completely dry, yielded much better results. Although the bot flicked some particles about and left some in the grouting between tiles, it cleared most of the mess on the first pass.

With most robovacs, the mop pad(s) can either be removed entirely or will lift to traverse hard floors, but there's no such option here. It isn't the end of the world – you just need to make sure you save the mopping for the last task of the day.

Mop performance

I was dismayed and confused to see there isn't a "mop only" mode here, especially since it looks like this bot can do both things separately ("Vacuum then mop" is an option). The mopping here is more involved than usual – the bot isn't just feeding water onto the mop pads, it's also scraping them clean as it goes. Perhaps for that reason, the S20 emits a rather grating, droning noise when it's mopping.

It also takes quite a while for the roller to completely soak through – in my test, I found it had almost completed mopping my whole (quite large) kitchen before the striped track marks became solid. It isn't unusual for hybrid robot vacuums to take a while to get their mops properly wet – I usually like to remove the pad and wet it down myself before setting any bot out on a mopping run. That's more of a hassle here, though, due to the way it's attached to the bot.

While mopping tests didn't go completely smoothly, I got good results in the end. First, I drizzled soy sauce and soy milk on a tiled floor, and asked the bot to do an area clean in Wet mode.

The first area clean was scuppered by the fact the bot seemed to identify the milk as an object to avoid. It skirted carefully around the puddle, smearing the soy sauce in the process, but not really achieving any actual cleaning.

After spreading the puddle of milk around a bit to blur the edges, it was more successful on the my second run (shown in the video below). It delivered on its promise of clearing the spillage rather than spreading it about, as can happen with flat mop pads.

Buoyed by my success, I also ran a test to see if it could clear smeared ketchup from the floor. Unfortunately, it didn't manage to make much of a difference to the mess after one pass in Wet mode. That's not unusual – it's a difficult test – but I had gone in with quite high hopes for the roller mop.

xxxx does the dock dry the roller successfully?

Navigation and object avoidance

Moving on to navigation. Again, my experience was a little hit-and-miss. The SwitchBot S20 generally followed a logical up-and-down pattern, but there were plenty of occasions where it decided to – in the words of Fleetwood Mac – go its own way. It also had a little trouble clearing taller room thresholds, but managed them all in the end.

To test object avoidance, I placed a tissue box, shoe and charging cable on the carpeted floor in my lounge, and asked the bot to conduct an area clean. As you might have suspected from the its behavior in the mopping test, the S20 certainly has some object avoidance skills.

It successfully avoided both the cable (unheard of) and the shoe, and after deciding it wouldn't fit between them, took a long route around the coffee table to complete cleaning on the other size of the requested zone. I've never seen that before. It wasn't quite perfect – it did push the tissue box right across the room – but it is still very solid performance.

• Performance score: 3.5 out of 5

SwitchBot S20 review: app

• Fairly easy to navigate
• Some frustrating quirks
• Part of a general SwitchBot app, for all the brand's products

The app is generally fine, with all the key adjustment options presented in a logical way. However, there are a few annoying quirks that stop it from being perfect.

I had a seriously frustrating time trying to divide the hall from the toilet, with the app refusing to give me the solid line required for an acceptable room division; it repeatedly bumped my adjuster off to elsewhere on the map. I managed to create a divider eventually, but it's on the slant, and not at right angles to any wall (to be clear, this is not the situation in real life).

Maps and rooms have an eight-character limit so you can't, for example, create a "Top floor". Oh, and there's no "Hall" or "Corridor" option within the room defaults – you can make one (just about, in the case of "corridor", and that pesky character limit) but I'm baffled as to why it isn't there in the first place.

Finally, because it's part of a general SwitchBot app for all the brand's products, you'll have to click through a couple of screens before you reach the robovac options. It's fine if you've bought into the SwitchBot ecosystem and have multiple products to control, but a bit annoying if you're using the robot vac on its own.

• App score: 3.5 out of 5

Should you buy the SwitchBot S20?

Buy it if...

Don't buy it if...

How I tested the SwitchBot S20

I used the SwitchBot S20 as my regular robot vacuum for a fortnight, to get a feel for how it performs day-to-day. I also ran a series of standardized vacuuming and mopping tests, to see how it would cope with spillages of all kinds. I dug into the app and assessed the different options there, as well as gauging how easy it would be to use and if things are presented in an intuitive way. I compared my findings to other robot vacuums I've tested, and weighed everything up against the cost to ascertain if the bot delivered good value for money.

Read more about how we test robot vacuum cleaners

• First reviewed October 2025

Ecovacs Deebot X8 Pro Omni: Two-minute review

I'll say one thing about Ecovacs Robotics: the brand might make too many robot vacuums that I, as a reviewer, can’t keep up with, but they all seem well thought out and catering to different needs and budgets. In 2024, it released a good-value robovac that I really liked in my Ecovacs Deebot T30 Omni review, and it’s impressed again with a higher-end model in the Deebot X8 Pro Omni reviewed here.

The X8 has plenty of smarts on board, with Ecovacs’ voice assistant Yiko now better at understanding commands and having a more conversational interaction than its previous iterations on older Deebots. Even if you don't interact with Yiko, the X8 has plenty of automated voice prompts that I very quickly found to be annoying – there's only so many times you can hear it say "don't worry" when its auto-emptying. While you can't switch these off, you can mute the volume entirely and depend on the app notifications to find out if the robot needs help at any time (like if it's stuck somewhere).

There's Google Assistant and Alexa support as well, but Yiko can handle a few extra commands that I found the other two voice assistants couldn't handle. To futureproof your smart home, the X8 Pro Omni also has Matter support, so you can control several smart devices via a single hub.

There's a camera on board that the bot uses to ‘see’ its surroundings for better navigation and obstacle avoidance, but it also provides a live feed that requires a passcode to be set up for access. There’s a Patrol mode here that turns the X8 Pro Omni bot into a security unit and it saves a bunch of photos for you to view whenever you like. You can also speak to your pets or kids if you need to check in on them (spy?) while you're at work. I’ve asked Ecovacs for clarification on how secure the camera is from hackers and will update this review as soon as I hear back.

These smart bells and whistles are all great, but this robot vacuum shines at its most important function: cleaning. It’s arguably one of the best automated cleaners I’ve tested in 2025, with edge cleaning its biggest asset thanks to an extending side brush and wet roller while vacuuming and mopping respectively. In fact, I’ve seen the X8 Pro Omni travel far closer to walls and furniture than many of its competitors, thus allowing its extendable parts to do their job effectively.

It puts its 18,000Pa of suction power to good use, with my carpets looking fresh and clean after a single run on its highest suction setting. It did a better job than what I saw in my Roborock Qrevo Edge review despite 18,500Pa of suction. And its roller mop was quite impressive compared to the mops pads I’d gotten used to previously, and that’s despite Ecovacs saying there’s no pressure applied by the roller on the floor.

There’s also a detergent dispenser on board and use of the soap is automatic provided you toggle it on in the app – Ecovacs says there’s always a ratio of 200:1 detergent and water mix in the pipeline to ensure the correct usage. Even without detergent I found my floors were left stain-free after every mop – in fact, it was better than I experienced in my Narwal Freo Z10 review that, at the time, I said had excellent mopping skills.

In four weeks of testing, the X8 Pro Omni got entangled in a cable just once because the side brush pulled a low-hanging wire towards the bumper, but otherwise its obstacle avoidance is very good. Even more impressive is its overall navigation – it invariably took the path of least resistance in my apartment, which meant it got its tasks done relatively quickly and, importantly, without consuming too much battery life.

Even the base station performance is top notch, with the onboard bin getting cleaned quite well each time it empties, and the wet-roller cleaning is impeccable. Water consumption is also lower than any other robovac I’ve tested before – I’ve previously had to refill the clean water after every full cleaning cycle in my small inner-city apartment, but this time I’ve had to do so once every four or five sessions (depending on its routine).

All in all, this was excellent performance from not just the robot vacuum, but I commend Ecovacs for streamlining its app experience as well. It’s much nicer to use, with plenty of customization options to suit your personal cleaning needs. Importantly, these custom routines are now very easy to access.

It's relatively noisier than some of its newer competitors, and it can't perform the climbing feats of the more expensive Dreame X50 Ultra, but if you value excellent cleaning performance above all else, I'd find it very easy to recommend the Deebot X8 Pro Omni – and it's often discounted.

Ecovacs Deebot X8 Pro Omni review: price & availability

• Available since March 2025
• List price: $1,099 / £1,099 / AU$2,499
• Incurs ongoing costs for dust bags and detergent

There are quite a few all-in-one robot vacuum cleaners at the same price point that Ecovacs Robotics is selling the X8 Pro Omni, but I think the Deebot can justify its premium cost on its feature set and performance.

It’s listed for $1,099 / £1,099 / AU$2,499 at full price but, at the time of writing, US and UK customers could pick it up for $899.99 / £999 respectively directly from the Ecovacs website. In Australia, it has dropped to AU$2,199 previously, but I suspect there’ll be better discounts around the world during major shopping events like Black Friday sales. It’s also available from select third-party retailers, including Amazon.

This price gets you not just excellent performance (explained below in detail), but you also get an impressive set of features, including voice assistant and Matter integration for a smarter home setup. With that in mind, I think the X8 Pro Omni is well priced compared to some of the other robovacs on the market, although Ecovacs has cheaper models that clean really well, as proved by the Deebot T30 Omni.

As an example of competitor pricing, the Narwal Freo Z Ultra cost $1,499/ AU$2,499 (unavailable in the UK) and is a good machine, but has 12,000Pa of suction compared to the Deebot’s 18,000Pa. Matching the suction power at 18,500Pa is the Roborock Qrevo Edge, but I found it hard to recommend at its list price of $1,599.99 / AU$2,799 (unavailable in the UK) based on its performance alone. On the other hand, Ecovacs' own Deebot T50 Max Pro Omni has 18,500Pa suction, the Yiko voice assistant and promises good cleaning at a lower price point of $799.99 / £799 / AU$1,799.

If you want more suction power, you’ll have to pay a lot more. For example, the Roborock Saros 10 with its whopping 22,000Pa of power will set you back $1,599.99 / £1,499.99 / AU$2,999 at full price, but is often discounted to match the X8's retail cost. The Dreame X50 Ultra Complete with 20,000Pa of suction is also expensive compared to the X8 Pro Omni, coming in at a list price of $1,699.99 / £1,299 / AU$2,999, but it can climb low steps that many of the newer machines can't.

As with any other hybrid robot vacuum, there is the ongoing cost of purchasing dust bags and detergent to take into consideration, although the latter is an optional add-on but I can see the need for it in homes with toddlers and pets. Unlike other brands, though, Ecovacs doesn’t supply a bottle of detergent in the box – at least not in Australia, where the testing for this review was conducted, due to “customs issues with international shipments” – so that’s $49.98 / £44 right off the bat for two 1L bottles in the US and UK (frequently discounted), but the same two-pack isn't available in Australia where a 1L bottle retails for AU$37.50.

The antibacterial dust bags come in packs of three and are available directly from Ecovacs Robotics for $24.99 / £17.99 / AU$29.90 at the time of writing.

Despite the ongoing cost, I think the Deebot X8 Pro Omni balances its feature set, performance and price well, but I’d recommend picking it up during a sale as it’s frequently discounted.

• Value score: 4 / 5

Ecovacs Deebot X8 Pro Omni review: specs

Ecovacs Deebot X8 Pro Omni review: Design

• Gold accents give it a classy look
• Extending side brush and wet roller mop
• Taller than the average robot but lacks a turret; tall but slim base station

I only had to take the robot out to see that Deebot X8 Pro Omni exudes class, thanks to some gold-colored accents on the otherwise black machine. Unlike the Narwal Freo Z10 that I reviewed prior to the X8 (which was a 11kg weight straight out of the box), the Deebot comes packed in layers, which makes it very easy to set up.

Robot design

Once you remove the wash tray from the top of the box, the next item to emerge is the X8 robot. The first thing that distinguishes it from many other robovacs is the lack of a top turret that typically houses the navigation tech. In the X8 Pro Omni, though, all the necessary sensors are on the front and the sides.

The lack of a turret means the X8 can roll under many low-lying pieces of furniture, although it's doesn't have the slimline form factor of the Roborock Saros 10R – it's a little taller, standing at 9.8cm/3.8in compared to the Saros at 8cm/3in, so there needs to be a minimum clearance of 10cm/4in for the Deebot to clean where many other robovacs can't.

A black plastic disc with the Ecovacs logo in gold covers the top of the robot, with a cutout to show off a gold three-pronged button. There's a tiny power icon in the center of the button to indicate that it's functional, but it's easy to miss and can appear purely decorative, although it lights up when in use and charging. Its functionality is listed on a large piece of cardboard inside the packaging: a short press to start or pause, a double press to extend or retract the wet roller, and a three-second long press to send the bot back to its dock. In the four weeks I spent with the machine, I never found the need to use the button – the robot does everything seamlessly.

Also under the top cover is a removable 220ml/7.4oz dustbin whose filter can be taken out for regular cleaning, plus a toggle switch for power, a Wi-Fi indicator light and a QR code to begin your app setup.

On the front of the robot are LiDAR sensors that it uses for mapping your home, while a camera also peeping through the same front window adds visual information that’s paired with artificial intelligence (AIVI) to navigate around obstacles. Along the sides are edge sensors that prevent it from banging into a wall or furniture, yet allowing the bot to get really close for effective cleaning. A mic and speaker are also housed within the front window.

The back of the robot has two charging points, but there’s also a water compartment inside with a clean capacity of 110ml and 95ml for dirty mop water. While you never have to worry about the water reservoir after setup, it can be released using an inset lever.

As with nearly every other Deebot I’ve tested (and many other brands besides), the underside of the robot houses the central bar brush that sports bristles arranged in a V shape to help it gather hair and fur without itself getting entangled. A single side brush towards the front of the bot rotates at varying speeds automatically depending on the size of the debris it senses – it slows down for larger pieces, similar to some of the high-end Roborocks. It can also extend out to reach into corners and clean along edges – both of which it manages to do better than many of its competitors because the robot itself gets really close to walls.

The rear of the undercarriage is dominated by a wet roller rather than circular mop pads which, according to Ecovacs, spins at a speed of 200rpm and has a scraper in its holder that constantly cleans out the dirty water as it rotates. 16 nozzles keep it supplied with clean water (and detergent if you’re using any).

The main wheels are quite large and, for the first time in any robot vacuum I’ve tested, I found them to be quite noisy when traveling over the tiles in my home. The sound was akin to very hard or brittle plastic moving over a rough surface, but in my time testing the X8 Pro Omni, I didn’t notice too much wear and tear on the wheels to account for the sound. A small omnidirectional caster wheel completes the robot’s physical setup.

Base station design

The base station itself isn’t much to write home about and is quite similar to several other models on the market. However, the gold accents on the lids of the two water tanks and branding on the front of the dock makes it look quite smart. The tanks have a 4L capacity each and are very easy to lift up, refill or clean out, then replace back on the dock.

Below the tanks is a covered nook that houses the detergent dispenser and a 3L dust bag. The dispenser is very easy to slide out and refill, while changing the dust bag is also very intuitive. It’s the cover for this nook that takes a little effort to remove each time you need access to either of its two inmates – it clips in very tightly indeed and you'll need to pull it with some force from the bottom of the panel.

Under the covered nook is the cavity for the robot, which sits on a washing tray. Inside the cavity, on the rear wall of the base station, you'll notice nozzles and charging points that keep the robot doing what it needs to do.

The base station is about as tall as many other brands offer, but it’s a little slimmer than, say, a Narwal dock. It will still need a decent amount of space to house, with enough clearance on the top to allow you to remove and replace the tanks.

• Design score: 4.5 / 5

Ecovacs Deebot X8 Pro Omni review: Performance

• Excellent vacuuming on high suction; top-notch mopping at most waterflow levels
• Arguably one of the best edge cleaners in the business
• Very good navigation and obstacle avoidance

The X8 Pro Omni for this review was tested in a 40sqm/430sqft inner-city apartment in Sydney, Australia, with the single bedroom fully carpeted, but the rest of the rooms featuring matte-finish tiles. The bathroom, however, has smoother tiles than the rest of the hard floors and it was part of the full map that the X8 had to clean. There are no stairs in the apartment to test cliff sensors, but there are a couple of low thresholds that were useful to test the X8 Pro Omni’s ability to clear some obstacles.

It’s also important for me to mention that I have no pets, but I’ve always found my bedroom carpet covered in my own (long) hair within 3-4 days of cleaning it, and is usually a very good test for any vacuum cleaner that I review.

Before starting the first cleaning session, I not only set up some custom routines (called Scenario Cleans in the Ecovacs Home app), but I also went through the advanced settings to toggle on some key features, including the carpet-first option as I have mixed flooring. This ensured that no matter what custom cleaning routine I might use, carpets were always vacuumed first to minimize the risk of them getting wet (more on the app experience later).

Overall performance of the X8 Pro Omni was excellent, but it’s important to note that it’s best used for the regular maintenance cleaning sessions that you might need on a daily or a weekly basis. It can’t replace the best cordless vacuum cleaners, but it definitely reduces the human effort and time spent on keeping your floors spick and span.

Vacuuming

Let’s start the performance section with the most basic function: vacuuming. I’ve already mentioned earlier in this review that the X8 Pro Omni was one of the best cleaning machines I’ve tested recently, exceeding the vacuuming performance I got from a Roborock and a Narwal that I tested prior to the Deebot. It really does put the 18,000Pa – which is in no way class-leading now – to very good use, particularly at its higher suction settings.

I usually have to set most robovacs to their maximum suction setting to get a decent clean on the bedroom carpet which, as I’ve described above, can get covered in my own hair within days. It’s been rare for me to see that carpet completely hair free after a single session even using the highest power level on other robovacs, but the X8 Pro Omni was been able to do just that each and every time it’s done a bedroom clean on its Max setting.

The Deebot’s got four suction settings and while I wasn’t particularly keen on the Quiet option which reduces the power significantly, I found the Standard and Strong settings to be perfect for hard floors, not once needing to step it up to Max on tiles.

The X8 Pro Omni has a suction-boost option which, as the name suggests, automatically increases power when it detects carpets, but I wasn’t able to notice the dynamic suction from sound alone when it went from hard floor to the mid-pile rug in my living room, although it seemed to do a very good job anyway. The audible change in suction was heard when it was vacuuming tiles in my kitchen, just under a cabinet – go figure.

To check its vacuuming abilities further, I conducted TechRadar’s standard tests of sprinkling tea dust and oats (representing two different-sized debris) on both hard floors and carpets. Doing a Zone clean and set at Strong suction, you can see from the video clip below that the X8 Pro Omni did a fantastic job on the tiles, even sucking up the tea dust from along the edge of the room.

If you look closely, you’ll also notice that the side brush slows down when it detects larger debris (oats) to make sure they all get pulled towards the bar brush.

On the carpet, which was also just a Zone clean done as a separate test, I set the robot to Max suction and found it cleared up the oats very well indeed, although tea dust that went deep into the fibers remained, and is visible in the video clip as a dark patch. I used my finger to loosen the carpet fibers and let the X8 run over it again, and voila, it was all cleaned up.

For both tests, I had the X8 Pro Omni set at moving in the Deep Clean pattern, which covers maximum floor space but time consuming, and I didn’t find the need to go over either of the spots with my Dyson V15s Detect Submarine.

Mopping

This is the first robot vacuum I’ve tested that features a wet roller rather than a pair of rotating mop pads and I really wasn’t sure what to expect. Needless to say, the X8 Pro Omni exceeded my expectations and then some with the way it tackled my hard floors. I will go so far as to say this was arguably the best mopping I’ve seen in a while by a robot vacuum.

From shaved Toblerone on the kitchen floor (I was making Toblerone cheesecake at the time) to a partially dried dollop of coriander-mint chutney, I was very impressed with how well it mopped up different (dry) spills.

I specify dry because you really don’t want a robot vacuum to clean a wet spill like cereal milk or curry sauce as everything in the undercarriage will get wet with the sticky liquid or the sauce, including the wheels and side brush. Cleaning the bot after that will be a chore in itself.

As with many other robot vacuums like itself, the X8 Pro Omni has three waterflow settings (Low, Medium and High) with a total of 50 levels. The lowest settings would be good for sensitive wood floors but, for me, I’ve always found higher waterflow rates better for my tiles, so I had the Deebot set at level 40 most of the time as I found 45 and 50 made my tiles very wet indeed, while 30 and lower barely left a damp streak.

My kitchen floors, in particular, need regular mopping and, given it’s right beside the dining room, this area of my apartment was a good test for the X8 Pro Omni. I found that in its Deep Cleaning speed and water level set to 45 but no detergent, there was a visible difference on my floor after its first run compared to what the Dyson V15s Detect Submarine had done prior to switching to the Deebot.

In fact, I’m not sure how much difference adding the detergent has done to my floors because it mopped up dusty footprints and dried, dusty water stains very well without any soap, and it did just as good a job with.

The stress test I put the X8 under was mopping up a partially dried dollop of coriander and mint chutney – it took a couple of passes at water level 40 with detergent in the dispenser to mop it up entirely, no stain left on the floor. The issue here was that the rotating side brush scattered a tiny bit of dry chutney away, but given it was set to Deep Clean, the bot moved over the entire zone twice (more on this below in the navigation section), running perpendicular to its initial path, which meant that scattered bit got mopped up later.

I enjoyed watching the extendable roller sneak in and out of the robot as soon as it sensed something on its side. This could be a wall or a furniture leg, but I found that it hardly left a dry spot.

On the product listing page on the Ecovacs website, the brand says there's no pressure applied by the wet roller on the floor, but given how well I saw it mop, I'd be surprised if this spec listing is correct. There's surely some pressure being applied for such effective mopping.

Another thing I love about the roller is that it's streak-free, particularly on the lower or mid waterflow rates. The only time I saw streaks (and not dirty ones, mind, just marks to show a mop has passed) was when the roller was saturated with clean water.

Navigation & obstacle avoidance

With the sole exception of one Zone cleaning session when it took the wrong path out of its dock to get to where it needed to in a different room, the Deebot X8 Pro Omni never once strayed from the path of least resistance.

It’s impressive navigation extends to doing full 360º rotations around furniture legs, and while the side brush didn’t always extend when I expected it to, the robot itself got so close to walls and furniture that I never had cause for complaint when it came to its navigation just millimeters from what is essentially an obstacle.

It’s three different “cleaning speeds” (or its movement pattern in a given space) are a little hit and miss if you really want a thorough job done. The best movement pattern I found was the Deep Cleaning, which means the robot cleans a space twice, first running in parallel lines just centimeters apart, then traveling perpendicular to ensure no spot is missed.

The other two cleaning speeds – Standard and Quick – aren’t bad, but because the parallel pathways are slightly more spread out compared to Deep Cleaning, I found that the occasional hair strand or speck of debris got missed. That said, “cleaning speed” as Ecovacs calls it is apt as these determine how quickly the X8 Pro Omni finishes its cleaning task.

Obstacle avoidance is also fantastic, as you can see from the video above when I conducted TechRadar’s standard test – creating an obstacle course of a slipper, shoe, a cable, socks and a tissue box. It avoided every single one of them without a hiccup, a feat I’ve not yet seen from any robot vacuum I’ve tested to date.

However, during one vacuum-only session in my bedroom, it got entangled in a USB-C cable that had one end dangling close to the floor and the side brush pulled it lower (see above image). I could hear Yiko complaining and a notification popped up on my phone that the X8 Pro Omni needed help. I disentangled it and it carried on with its work after I used a voice command to “restart cleaning”.

During a full-house clean, I shut a door to one room to see how it would react to its map not quite being accurate and it had no issues. It got close to the door, pivoted a little on an axis to see if it could spot an opening, but carried on doing what it needed to in the rest of the apartment without any complaints.

Base station performance

Whether it’s sucking out all the dirt and debris from the onboard dustbin or cleaning the wet roller, the X8 Pro Omni base station performance is so good that it justifies the high price tag this machine demands.

Only the tiniest amount of dirt remains inside the onboard dustbin after each vacuuming run that you, honestly, never have to worry about. The only time you’ll need to remove the small onboard bin is when the filter needs cleaning or replacing, or if something gets stuck inside that requires manual removal (never once happened during my testing).

The dustbag inside the base station has a 3L capacity, so it will last you a long while, albeit that will depend on how often you clean, how large your home is and whether you have pets.

Dispensing detergent happened as required, as did filling up the robot’s onboard 110ml clean water compartment. Even pumping out dirty water from the 90ml partition worked well during my testing – you can remove this compartment to clean out the filter inside.

The washing of the roller was impeccable. It looked nearly new after my running it a few times a week for four weeks and the shortest two-hour hot-air drying is more than enough to ensure there’s no trace of dampness that could encourage bacterial or fungal growth. In fact, at the end of my four-week trial period, I found the roller soft to the touch as it was on day one.

The wash tray can also be cleaned out automatically by the machine from time to time, so the only hands-on task you’ll have to do regularly is clean out the dirty water and refill the fresh water. Every couple of months or so, you’ll also need to swap out the dust bag or refill the detergent dispenser if you’ve been using soap for mopping. The ‘health’ of all the accessories and parts is listed in the app, so you’ll know when to do what anyway.

Battery life

Ecovacs says there’s a 6,400mAh battery pack powering the X8 Pro Omni robot, which is rated for up to 228 minutes or a maximum area coverage of 240sqm / 2,580sq ft. In my testing – which admittedly was done in a much smaller space – I found that a number like that is nigh impossible to achieve unless you run the robot at the lowest settings possible, which I doubt anyone will want to do.

In Standard suction and medium waterflow levels during a full 40sqm clean doing a single vacuum-and-mop pass, the X8’s battery dropped to 20%. Bumping it up to Max suction and high waterflow rates, I found the robot needed to go back for a top up at the mains after completing five out of the six rooms in its map – about 78 minutes. Both these tests were conducted at the Deep Clean speed, so more time consuming.

Extrapolating that for Standard or Quiet cleans, you will be able to run the X8 Pro Omni for around 150 minutes, but keep in mind that it boasts dynamic suction as well, so total battery life will depend on how many carpets it encounters or how tough the cleaning job is.

That’s not bad battery life, but it’s not class-leading as models like the Narwal Freo Z10 and Dreame X40 Ultra can outdo the X8. Larger homes will find that a full cleaning task may well require recharging to complete. That said, the X8 is smart enough to learn how much battery it needs to finish a job, so a full-home clean may not necessarily take all day.

• Performance score: 4.5 / 5

Ecovacs Deebot X8 Pro Omni review: app control

• Improved Ecovacs Home app
• More streamlined setup for robot and base station settings
• Map might need extensive editing, but is easy to do

I’ve previously not been very impressed with the Ecovacs Home app as it wasn’t as streamlined as I would have liked, but that’s changed now – the updated version of the app that I used with the X8 Pro Omni is a significant improvement.

It’s now easier to access the custom cleaning routines – renamed Scenario Cleans from just plain ol’ Scenarios – as they're listed on the homepage of the device, and each of these are also very easy to customize and set up.

To access more settings options, the Ecovacs Home app has always allowed you to ‘Enter’ a more detailed page with your home’s map laid out, but now the robot and base station advanced settings are right there on that same screen – the robot stuff is right at the bottom while switching tabs gives you the base station setup. This is a much better app experience than I encountered only a few months ago from the time of publication of this review.

As I’ve already mentioned, there’s the Yiko voice assistant on board and, in all honesty, I really don’t see the need for verbal communication with the device, but my inner lazy self took a shine to it for a very short time. All routines and settings work so well, that Yiko doesn’t need to be there, but it was nice to see that it can understand commands better than before.

During my testing, it didn’t comprehend its own name (see the screenshot below where it hears its name as "you go"), but that didn’t stop it from performing the task I asked it to do (pause cleaning). I immediately asked it to restart again – no problem whatsoever.

You’ll also hear a lot of oral prompts every time the robot has to perform a task. This can quickly get very annoying, particularly when the self-empty of the bin takes place because it’s a long, drawn-out message telling you “don’t worry” about the loud noises. There’s sadly no way to turn these off, but you can drop the volume right down to ‘mute’, but this also means you won’t hear when the bot needs help. Thankfully there are app notifications to help with that. The volume is also associated with a ‘ding’ sound the robot makes every time you adjust a setting – these too will be muted if you turn the volume slider right down.

Every setting on the app is well explained and easy to understand in case you want to use it, but I would advise that you take some time to go through all options before you begin using the X8 Pro Omni so that you get the best results possible.

For example, you can set and forget the cleaning sequence for the whole house and, no matter which rooms you choose for whatever custom routine you might have, the X8 Pro Omni will follow that. This can be edited at any time too. As I’ve mentioned before, you can also toggle on carpet-first cleaning in the robot settings, set how you want the auto-empty to take place or how often you want the wet roller washed. You even get control over how long you want the base station to spend drying it with hot air. You’ll also need to toggle on auto detergent use, otherwise the base station won’t use soap even if you’ve filled the dispenser.

There is a lot of control here, although it would be nice to see Ecovacs allow the user to choose how many passes the X8 will do when on an automatic AI-powered cleaning run (it’s always just one in this mode).

• App score: 4 / 5

Should I buy the Ecovacs Deebot X8 Pro Omni?

Don't buy it if...

Ecovacs Deebot X8 Pro Omni review: Also consider

How I tested the Ecovacs Deebot X8 Pro Omni

• Used three to four times a week for four weeks
• Tried various suction and waterflow levels in a 40sqm test space with mixed flooring
• Experimented with the Yiko voice assistant and tweaked various advanced settings

I received my review unit of the Ecovacs Deebot X8 Pro Omni in the third of week of July 2025 and spent the next four weeks running it on various settings and custom routines mostly three times a week, but it has done more sessions during the last week when I conducted my obstacle course test.

I found that once all the settings were as I'd want them, the X8 Pro Omni just did what it needed to do without any fuss. During that time, I found it didn't use as much clean water as other robot vacuums (especially compared to the Narwal Freo Z10 I tested prior to it) and, hence, found I didn't need to keep refilling the clean-water tank as much. Of course, that's also because the test space is small (just 40sqm/430sqft) and has mixed flooring with carpet in the bedroom and a large rug in the living room.

I set up four different custom routines, or Shortcut Cleans as the app labels them, and tested each one 2-3 times over the course of the four-week testing period. I also ran single-room and zone cleaning sessions to see how the X8 Pro Omni performs in its auto-cleaning mode.

While I began testing the X8 without any detergent for the first week, I added a small quantity to the dispenser on week 2 of testing.

I put it through TechRadar's usual tests for vacuuming and mopping (using oats and tea from a bag for the former, and a dollop of chutney for the latter). I also conducted our standard obstacle avoidance test, but threw in two types of socks and a tissue box for good measure.

As you've probably already read, the X8 Pro Omni passed all its tests with flying colors, with only the one hiccup when it got entangled in a low-hanging cable during a vacuum-only session.

Read more about how we test vacuum cleaners

[First reviewed August 2025]

Ecovacs Deebot X8 Pro Omni: Two-minute review

I'll say one thing about Ecovacs Robotics: the brand might make too many robot vacuums that I, as a reviewer, can’t keep up with, but they all seem well thought out and catering to different needs and budgets. In 2024, it released a good-value robovac that I really liked in my Ecovacs Deebot T30 Omni review, and it’s impressed again with a higher-end model in the Deebot X8 Pro Omni reviewed here.

The X8 has plenty of smarts on board, with Ecovacs’ voice assistant Yiko now better at understanding commands and having a more conversational interaction than its previous iterations on older Deebots. Even if you don't interact with Yiko, the X8 has plenty of automated voice prompts that I very quickly found to be annoying – there's only so many times you can hear it say "don't worry" when its auto-emptying. While you can't switch these off, you can mute the volume entirely and depend on the app notifications to find out if the robot needs help at any time (like if it's stuck somewhere).

There's Google Assistant and Alexa support as well, but Yiko can handle a few extra commands that I found the other two voice assistants couldn't handle. To futureproof your smart home, the X8 Pro Omni also has Matter support, so you can control several smart devices via a single hub.

There's a camera on board that the bot uses to ‘see’ its surroundings for better navigation and obstacle avoidance, but it also provides a live feed that requires a passcode to be set up for access. There’s a Patrol mode here that turns the X8 Pro Omni bot into a security unit and it saves a bunch of photos for you to view whenever you like. You can also speak to your pets or kids if you need to check in on them (spy?) while you're at work. I’ve asked Ecovacs for clarification on how secure the camera is from hackers and will update this review as soon as I hear back.

These smart bells and whistles are all great, but this robot vacuum shines at its most important function: cleaning. It’s arguably one of the best automated cleaners I’ve tested in 2025, with edge cleaning its biggest asset thanks to an extending side brush and wet roller while vacuuming and mopping respectively. In fact, I’ve seen the X8 Pro Omni travel far closer to walls and furniture than many of its competitors, thus allowing its extendable parts to do their job effectively.

It puts its 18,000Pa of suction power to good use, with my carpets looking fresh and clean after a single run on its highest suction setting. It did a better job than what I saw in my Roborock Qrevo Edge review despite 18,500Pa of suction. And its roller mop was quite impressive compared to the mops pads I’d gotten used to previously, and that’s despite Ecovacs saying there’s no pressure applied by the roller on the floor.

There’s also a detergent dispenser on board and use of the soap is automatic provided you toggle it on in the app – Ecovacs says there’s always a ratio of 200:1 detergent and water mix in the pipeline to ensure the correct usage. Even without detergent I found my floors were left stain-free after every mop – in fact, it was better than I experienced in my Narwal Freo Z10 review that, at the time, I said had excellent mopping skills.

In four weeks of testing, the X8 Pro Omni got entangled in a cable just once because the side brush pulled a low-hanging wire towards the bumper, but otherwise its obstacle avoidance is very good. Even more impressive is its overall navigation – it invariably took the path of least resistance in my apartment, which meant it got its tasks done relatively quickly and, importantly, without consuming too much battery life.

Even the base station performance is top notch, with the onboard bin getting cleaned quite well each time it empties, and the wet-roller cleaning is impeccable. Water consumption is also lower than any other robovac I’ve tested before – I’ve previously had to refill the clean water after every full cleaning cycle in my small inner-city apartment, but this time I’ve had to do so once every four or five sessions (depending on its routine).

All in all, this was excellent performance from not just the robot vacuum, but I commend Ecovacs for streamlining its app experience as well. It’s much nicer to use, with plenty of customization options to suit your personal cleaning needs. Importantly, these custom routines are now very easy to access.

It's relatively noisier than some of its newer competitors, and it can't perform the climbing feats of the more expensive Dreame X50 Ultra, but if you value excellent cleaning performance above all else, I'd find it very easy to recommend the Deebot X8 Pro Omni – and it's often discounted.

Ecovacs Deebot X8 Pro Omni review: price & availability

• Available since March 2025
• List price: $1,099 / £1,099 / AU$2,499
• Incurs ongoing costs for dust bags and detergent

There are quite a few all-in-one robot vacuum cleaners at the same price point that Ecovacs Robotics is selling the X8 Pro Omni, but I think the Deebot can justify its premium cost on its feature set and performance.

It’s listed for $1,099 / £1,099 / AU$2,499 at full price but, at the time of writing, US and UK customers could pick it up for $899.99 / £999 respectively directly from the Ecovacs website. In Australia, it has dropped to AU$2,199 previously, but I suspect there’ll be better discounts around the world during major shopping events like Black Friday sales. It’s also available from select third-party retailers, including Amazon.

This price gets you not just excellent performance (explained below in detail), but you also get an impressive set of features, including voice assistant and Matter integration for a smarter home setup. With that in mind, I think the X8 Pro Omni is well priced compared to some of the other robovacs on the market, although Ecovacs has cheaper models that clean really well, as proved by the Deebot T30 Omni.

As an example of competitor pricing, the Narwal Freo Z Ultra cost $1,499/ AU$2,499 (unavailable in the UK) and is a good machine, but has 12,000Pa of suction compared to the Deebot’s 18,000Pa. Matching the suction power at 18,500Pa is the Roborock Qrevo Edge, but I found it hard to recommend at its list price of $1,599.99 / AU$2,799 (unavailable in the UK) based on its performance alone. On the other hand, Ecovacs' own Deebot T50 Max Pro Omni has 18,500Pa suction, the Yiko voice assistant and promises good cleaning at a lower price point of $799.99 / £799 / AU$1,799.

If you want more suction power, you’ll have to pay a lot more. For example, the Roborock Saros 10 with its whopping 22,000Pa of power will set you back $1,599.99 / £1,499.99 / AU$2,999 at full price, but is often discounted to match the X8's retail cost. The Dreame X50 Ultra Complete with 20,000Pa of suction is also expensive compared to the X8 Pro Omni, coming in at a list price of $1,699.99 / £1,299 / AU$2,999, but it can climb low steps that many of the newer machines can't.

As with any other hybrid robot vacuum, there is the ongoing cost of purchasing dust bags and detergent to take into consideration, although the latter is an optional add-on but I can see the need for it in homes with toddlers and pets. Unlike other brands, though, Ecovacs doesn’t supply a bottle of detergent in the box – at least not in Australia, where the testing for this review was conducted, due to “customs issues with international shipments” – so that’s $49.98 / £44 right off the bat for two 1L bottles in the US and UK (frequently discounted), but the same two-pack isn't available in Australia where a 1L bottle retails for AU$37.50.

The antibacterial dust bags come in packs of three and are available directly from Ecovacs Robotics for $24.99 / £17.99 / AU$29.90 at the time of writing.

Despite the ongoing cost, I think the Deebot X8 Pro Omni balances its feature set, performance and price well, but I’d recommend picking it up during a sale as it’s frequently discounted.

• Value score: 4 / 5

Ecovacs Deebot X8 Pro Omni review: specs

Ecovacs Deebot X8 Pro Omni review: Design

• Gold accents give it a classy look
• Extending side brush and wet roller mop
• Taller than the average robot but lacks a turret; tall but slim base station

I only had to take the robot out to see that Deebot X8 Pro Omni exudes class, thanks to some gold-colored accents on the otherwise black machine. Unlike the Narwal Freo Z10 that I reviewed prior to the X8 (which was a 11kg weight straight out of the box), the Deebot comes packed in layers, which makes it very easy to set up.

Robot design

Once you remove the wash tray from the top of the box, the next item to emerge is the X8 robot. The first thing that distinguishes it from many other robovacs is the lack of a top turret that typically houses the navigation tech. In the X8 Pro Omni, though, all the necessary sensors are on the front and the sides.

The lack of a turret means the X8 can roll under many low-lying pieces of furniture, although it's doesn't have the slimline form factor of the Roborock Saros 10R – it's a little taller, standing at 9.8cm/3.8in compared to the Saros at 8cm/3in, so there needs to be a minimum clearance of 10cm/4in for the Deebot to clean where many other robovacs can't.

A black plastic disc with the Ecovacs logo in gold covers the top of the robot, with a cutout to show off a gold three-pronged button. There's a tiny power icon in the center of the button to indicate that it's functional, but it's easy to miss and can appear purely decorative, although it lights up when in use and charging. Its functionality is listed on a large piece of cardboard inside the packaging: a short press to start or pause, a double press to extend or retract the wet roller, and a three-second long press to send the bot back to its dock. In the four weeks I spent with the machine, I never found the need to use the button – the robot does everything seamlessly.

Also under the top cover is a removable 220ml/7.4oz dustbin whose filter can be taken out for regular cleaning, plus a toggle switch for power, a Wi-Fi indicator light and a QR code to begin your app setup.

On the front of the robot are LiDAR sensors that it uses for mapping your home, while a camera also peeping through the same front window adds visual information that’s paired with artificial intelligence (AIVI) to navigate around obstacles. Along the sides are edge sensors that prevent it from banging into a wall or furniture, yet allowing the bot to get really close for effective cleaning. A mic and speaker are also housed within the front window.

The back of the robot has two charging points, but there’s also a water compartment inside with a clean capacity of 110ml and 95ml for dirty mop water. While you never have to worry about the water reservoir after setup, it can be released using an inset lever.

As with nearly every other Deebot I’ve tested (and many other brands besides), the underside of the robot houses the central bar brush that sports bristles arranged in a V shape to help it gather hair and fur without itself getting entangled. A single side brush towards the front of the bot rotates at varying speeds automatically depending on the size of the debris it senses – it slows down for larger pieces, similar to some of the high-end Roborocks. It can also extend out to reach into corners and clean along edges – both of which it manages to do better than many of its competitors because the robot itself gets really close to walls.

The rear of the undercarriage is dominated by a wet roller rather than circular mop pads which, according to Ecovacs, spins at a speed of 200rpm and has a scraper in its holder that constantly cleans out the dirty water as it rotates. 16 nozzles keep it supplied with clean water (and detergent if you’re using any).

The main wheels are quite large and, for the first time in any robot vacuum I’ve tested, I found them to be quite noisy when traveling over the tiles in my home. The sound was akin to very hard or brittle plastic moving over a rough surface, but in my time testing the X8 Pro Omni, I didn’t notice too much wear and tear on the wheels to account for the sound. A small omnidirectional caster wheel completes the robot’s physical setup.

Base station design

The base station itself isn’t much to write home about and is quite similar to several other models on the market. However, the gold accents on the lids of the two water tanks and branding on the front of the dock makes it look quite smart. The tanks have a 4L capacity each and are very easy to lift up, refill or clean out, then replace back on the dock.

Below the tanks is a covered nook that houses the detergent dispenser and a 3L dust bag. The dispenser is very easy to slide out and refill, while changing the dust bag is also very intuitive. It’s the cover for this nook that takes a little effort to remove each time you need access to either of its two inmates – it clips in very tightly indeed and you'll need to pull it with some force from the bottom of the panel.

Under the covered nook is the cavity for the robot, which sits on a washing tray. Inside the cavity, on the rear wall of the base station, you'll notice nozzles and charging points that keep the robot doing what it needs to do.

The base station is about as tall as many other brands offer, but it’s a little slimmer than, say, a Narwal dock. It will still need a decent amount of space to house, with enough clearance on the top to allow you to remove and replace the tanks.

• Design score: 4.5 / 5

Ecovacs Deebot X8 Pro Omni review: Performance

• Excellent vacuuming on high suction; top-notch mopping at most waterflow levels
• Arguably one of the best edge cleaners in the business
• Very good navigation and obstacle avoidance

The X8 Pro Omni for this review was tested in a 40sqm/430sqft inner-city apartment in Sydney, Australia, with the single bedroom fully carpeted, but the rest of the rooms featuring matte-finish tiles. The bathroom, however, has smoother tiles than the rest of the hard floors and it was part of the full map that the X8 had to clean. There are no stairs in the apartment to test cliff sensors, but there are a couple of low thresholds that were useful to test the X8 Pro Omni’s ability to clear some obstacles.

It’s also important for me to mention that I have no pets, but I’ve always found my bedroom carpet covered in my own (long) hair within 3-4 days of cleaning it, and is usually a very good test for any vacuum cleaner that I review.

Before starting the first cleaning session, I not only set up some custom routines (called Scenario Cleans in the Ecovacs Home app), but I also went through the advanced settings to toggle on some key features, including the carpet-first option as I have mixed flooring. This ensured that no matter what custom cleaning routine I might use, carpets were always vacuumed first to minimize the risk of them getting wet (more on the app experience later).

Overall performance of the X8 Pro Omni was excellent, but it’s important to note that it’s best used for the regular maintenance cleaning sessions that you might need on a daily or a weekly basis. It can’t replace the best cordless vacuum cleaners, but it definitely reduces the human effort and time spent on keeping your floors spick and span.

Vacuuming

Let’s start the performance section with the most basic function: vacuuming. I’ve already mentioned earlier in this review that the X8 Pro Omni was one of the best cleaning machines I’ve tested recently, exceeding the vacuuming performance I got from a Roborock and a Narwal that I tested prior to the Deebot. It really does put the 18,000Pa – which is in no way class-leading now – to very good use, particularly at its higher suction settings.

I usually have to set most robovacs to their maximum suction setting to get a decent clean on the bedroom carpet which, as I’ve described above, can get covered in my own hair within days. It’s been rare for me to see that carpet completely hair free after a single session even using the highest power level on other robovacs, but the X8 Pro Omni was been able to do just that each and every time it’s done a bedroom clean on its Max setting.

The Deebot’s got four suction settings and while I wasn’t particularly keen on the Quiet option which reduces the power significantly, I found the Standard and Strong settings to be perfect for hard floors, not once needing to step it up to Max on tiles.

The X8 Pro Omni has a suction-boost option which, as the name suggests, automatically increases power when it detects carpets, but I wasn’t able to notice the dynamic suction from sound alone when it went from hard floor to the mid-pile rug in my living room, although it seemed to do a very good job anyway. The audible change in suction was heard when it was vacuuming tiles in my kitchen, just under a cabinet – go figure.

To check its vacuuming abilities further, I conducted TechRadar’s standard tests of sprinkling tea dust and oats (representing two different-sized debris) on both hard floors and carpets. Doing a Zone clean and set at Strong suction, you can see from the video clip below that the X8 Pro Omni did a fantastic job on the tiles, even sucking up the tea dust from along the edge of the room.

If you look closely, you’ll also notice that the side brush slows down when it detects larger debris (oats) to make sure they all get pulled towards the bar brush.

On the carpet, which was also just a Zone clean done as a separate test, I set the robot to Max suction and found it cleared up the oats very well indeed, although tea dust that went deep into the fibers remained, and is visible in the video clip as a dark patch. I used my finger to loosen the carpet fibers and let the X8 run over it again, and voila, it was all cleaned up.

For both tests, I had the X8 Pro Omni set at moving in the Deep Clean pattern, which covers maximum floor space but time consuming, and I didn’t find the need to go over either of the spots with my Dyson V15s Detect Submarine.

Mopping

This is the first robot vacuum I’ve tested that features a wet roller rather than a pair of rotating mop pads and I really wasn’t sure what to expect. Needless to say, the X8 Pro Omni exceeded my expectations and then some with the way it tackled my hard floors. I will go so far as to say this was arguably the best mopping I’ve seen in a while by a robot vacuum.

From shaved Toblerone on the kitchen floor (I was making Toblerone cheesecake at the time) to a partially dried dollop of coriander-mint chutney, I was very impressed with how well it mopped up different (dry) spills.

I specify dry because you really don’t want a robot vacuum to clean a wet spill like cereal milk or curry sauce as everything in the undercarriage will get wet with the sticky liquid or the sauce, including the wheels and side brush. Cleaning the bot after that will be a chore in itself.

As with many other robot vacuums like itself, the X8 Pro Omni has three waterflow settings (Low, Medium and High) with a total of 50 levels. The lowest settings would be good for sensitive wood floors but, for me, I’ve always found higher waterflow rates better for my tiles, so I had the Deebot set at level 40 most of the time as I found 45 and 50 made my tiles very wet indeed, while 30 and lower barely left a damp streak.

My kitchen floors, in particular, need regular mopping and, given it’s right beside the dining room, this area of my apartment was a good test for the X8 Pro Omni. I found that in its Deep Cleaning speed and water level set to 45 but no detergent, there was a visible difference on my floor after its first run compared to what the Dyson V15s Detect Submarine had done prior to switching to the Deebot.

In fact, I’m not sure how much difference adding the detergent has done to my floors because it mopped up dusty footprints and dried, dusty water stains very well without any soap, and it did just as good a job with.

The stress test I put the X8 under was mopping up a partially dried dollop of coriander and mint chutney – it took a couple of passes at water level 40 with detergent in the dispenser to mop it up entirely, no stain left on the floor. The issue here was that the rotating side brush scattered a tiny bit of dry chutney away, but given it was set to Deep Clean, the bot moved over the entire zone twice (more on this below in the navigation section), running perpendicular to its initial path, which meant that scattered bit got mopped up later.

I enjoyed watching the extendable roller sneak in and out of the robot as soon as it sensed something on its side. This could be a wall or a furniture leg, but I found that it hardly left a dry spot.

On the product listing page on the Ecovacs website, the brand says there's no pressure applied by the wet roller on the floor, but given how well I saw it mop, I'd be surprised if this spec listing is correct. There's surely some pressure being applied for such effective mopping.

Another thing I love about the roller is that it's streak-free, particularly on the lower or mid waterflow rates. The only time I saw streaks (and not dirty ones, mind, just marks to show a mop has passed) was when the roller was saturated with clean water.

Navigation & obstacle avoidance

With the sole exception of one Zone cleaning session when it took the wrong path out of its dock to get to where it needed to in a different room, the Deebot X8 Pro Omni never once strayed from the path of least resistance.

It’s impressive navigation extends to doing full 360º rotations around furniture legs, and while the side brush didn’t always extend when I expected it to, the robot itself got so close to walls and furniture that I never had cause for complaint when it came to its navigation just millimeters from what is essentially an obstacle.

It’s three different “cleaning speeds” (or its movement pattern in a given space) are a little hit and miss if you really want a thorough job done. The best movement pattern I found was the Deep Cleaning, which means the robot cleans a space twice, first running in parallel lines just centimeters apart, then traveling perpendicular to ensure no spot is missed.

The other two cleaning speeds – Standard and Quick – aren’t bad, but because the parallel pathways are slightly more spread out compared to Deep Cleaning, I found that the occasional hair strand or speck of debris got missed. That said, “cleaning speed” as Ecovacs calls it is apt as these determine how quickly the X8 Pro Omni finishes its cleaning task.

Obstacle avoidance is also fantastic, as you can see from the video above when I conducted TechRadar’s standard test – creating an obstacle course of a slipper, shoe, a cable, socks and a tissue box. It avoided every single one of them without a hiccup, a feat I’ve not yet seen from any robot vacuum I’ve tested to date.

However, during one vacuum-only session in my bedroom, it got entangled in a USB-C cable that had one end dangling close to the floor and the side brush pulled it lower (see above image). I could hear Yiko complaining and a notification popped up on my phone that the X8 Pro Omni needed help. I disentangled it and it carried on with its work after I used a voice command to “restart cleaning”.

During a full-house clean, I shut a door to one room to see how it would react to its map not quite being accurate and it had no issues. It got close to the door, pivoted a little on an axis to see if it could spot an opening, but carried on doing what it needed to in the rest of the apartment without any complaints.

Base station performance

Whether it’s sucking out all the dirt and debris from the onboard dustbin or cleaning the wet roller, the X8 Pro Omni base station performance is so good that it justifies the high price tag this machine demands.

Only the tiniest amount of dirt remains inside the onboard dustbin after each vacuuming run that you, honestly, never have to worry about. The only time you’ll need to remove the small onboard bin is when the filter needs cleaning or replacing, or if something gets stuck inside that requires manual removal (never once happened during my testing).

The dustbag inside the base station has a 3L capacity, so it will last you a long while, albeit that will depend on how often you clean, how large your home is and whether you have pets.

Dispensing detergent happened as required, as did filling up the robot’s onboard 110ml clean water compartment. Even pumping out dirty water from the 90ml partition worked well during my testing – you can remove this compartment to clean out the filter inside.

The washing of the roller was impeccable. It looked nearly new after my running it a few times a week for four weeks and the shortest two-hour hot-air drying is more than enough to ensure there’s no trace of dampness that could encourage bacterial or fungal growth. In fact, at the end of my four-week trial period, I found the roller soft to the touch as it was on day one.

The wash tray can also be cleaned out automatically by the machine from time to time, so the only hands-on task you’ll have to do regularly is clean out the dirty water and refill the fresh water. Every couple of months or so, you’ll also need to swap out the dust bag or refill the detergent dispenser if you’ve been using soap for mopping. The ‘health’ of all the accessories and parts is listed in the app, so you’ll know when to do what anyway.

Battery life

Ecovacs says there’s a 6,400mAh battery pack powering the X8 Pro Omni robot, which is rated for up to 228 minutes or a maximum area coverage of 240sqm / 2,580sq ft. In my testing – which admittedly was done in a much smaller space – I found that a number like that is nigh impossible to achieve unless you run the robot at the lowest settings possible, which I doubt anyone will want to do.

In Standard suction and medium waterflow levels during a full 40sqm clean doing a single vacuum-and-mop pass, the X8’s battery dropped to 20%. Bumping it up to Max suction and high waterflow rates, I found the robot needed to go back for a top up at the mains after completing five out of the six rooms in its map – about 78 minutes. Both these tests were conducted at the Deep Clean speed, so more time consuming.

Extrapolating that for Standard or Quiet cleans, you will be able to run the X8 Pro Omni for around 150 minutes, but keep in mind that it boasts dynamic suction as well, so total battery life will depend on how many carpets it encounters or how tough the cleaning job is.

That’s not bad battery life, but it’s not class-leading as models like the Narwal Freo Z10 and Dreame X40 Ultra can outdo the X8. Larger homes will find that a full cleaning task may well require recharging to complete. That said, the X8 is smart enough to learn how much battery it needs to finish a job, so a full-home clean may not necessarily take all day.

• Performance score: 4.5 / 5

Ecovacs Deebot X8 Pro Omni review: app control

• Improved Ecovacs Home app
• More streamlined setup for robot and base station settings
• Map might need extensive editing, but is easy to do

I’ve previously not been very impressed with the Ecovacs Home app as it wasn’t as streamlined as I would have liked, but that’s changed now – the updated version of the app that I used with the X8 Pro Omni is a significant improvement.

It’s now easier to access the custom cleaning routines – renamed Scenario Cleans from just plain ol’ Scenarios – as they're listed on the homepage of the device, and each of these are also very easy to customize and set up.

To access more settings options, the Ecovacs Home app has always allowed you to ‘Enter’ a more detailed page with your home’s map laid out, but now the robot and base station advanced settings are right there on that same screen – the robot stuff is right at the bottom while switching tabs gives you the base station setup. This is a much better app experience than I encountered only a few months ago from the time of publication of this review.

As I’ve already mentioned, there’s the Yiko voice assistant on board and, in all honesty, I really don’t see the need for verbal communication with the device, but my inner lazy self took a shine to it for a very short time. All routines and settings work so well, that Yiko doesn’t need to be there, but it was nice to see that it can understand commands better than before.

During my testing, it didn’t comprehend its own name (see the screenshot below where it hears its name as "you go"), but that didn’t stop it from performing the task I asked it to do (pause cleaning). I immediately asked it to restart again – no problem whatsoever.

You’ll also hear a lot of oral prompts every time the robot has to perform a task. This can quickly get very annoying, particularly when the self-empty of the bin takes place because it’s a long, drawn-out message telling you “don’t worry” about the loud noises. There’s sadly no way to turn these off, but you can drop the volume right down to ‘mute’, but this also means you won’t hear when the bot needs help. Thankfully there are app notifications to help with that. The volume is also associated with a ‘ding’ sound the robot makes every time you adjust a setting – these too will be muted if you turn the volume slider right down.

Every setting on the app is well explained and easy to understand in case you want to use it, but I would advise that you take some time to go through all options before you begin using the X8 Pro Omni so that you get the best results possible.

For example, you can set and forget the cleaning sequence for the whole house and, no matter which rooms you choose for whatever custom routine you might have, the X8 Pro Omni will follow that. This can be edited at any time too. As I’ve mentioned before, you can also toggle on carpet-first cleaning in the robot settings, set how you want the auto-empty to take place or how often you want the wet roller washed. You even get control over how long you want the base station to spend drying it with hot air. You’ll also need to toggle on auto detergent use, otherwise the base station won’t use soap even if you’ve filled the dispenser.

There is a lot of control here, although it would be nice to see Ecovacs allow the user to choose how many passes the X8 will do when on an automatic AI-powered cleaning run (it’s always just one in this mode).

• App score: 4 / 5

Should I buy the Ecovacs Deebot X8 Pro Omni?

Don't buy it if...

Ecovacs Deebot X8 Pro Omni review: Also consider

How I tested the Ecovacs Deebot X8 Pro Omni

• Used three to four times a week for four weeks
• Tried various suction and waterflow levels in a 40sqm test space with mixed flooring
• Experimented with the Yiko voice assistant and tweaked various advanced settings

I received my review unit of the Ecovacs Deebot X8 Pro Omni in the third of week of July 2025 and spent the next four weeks running it on various settings and custom routines mostly three times a week, but it has done more sessions during the last week when I conducted my obstacle course test.

I found that once all the settings were as I'd want them, the X8 Pro Omni just did what it needed to do without any fuss. During that time, I found it didn't use as much clean water as other robot vacuums (especially compared to the Narwal Freo Z10 I tested prior to it) and, hence, found I didn't need to keep refilling the clean-water tank as much. Of course, that's also because the test space is small (just 40sqm/430sqft) and has mixed flooring with carpet in the bedroom and a large rug in the living room.

I set up four different custom routines, or Shortcut Cleans as the app labels them, and tested each one 2-3 times over the course of the four-week testing period. I also ran single-room and zone cleaning sessions to see how the X8 Pro Omni performs in its auto-cleaning mode.

While I began testing the X8 without any detergent for the first week, I added a small quantity to the dispenser on week 2 of testing.

I put it through TechRadar's usual tests for vacuuming and mopping (using oats and tea from a bag for the former, and a dollop of chutney for the latter). I also conducted our standard obstacle avoidance test, but threw in two types of socks and a tissue box for good measure.

As you've probably already read, the X8 Pro Omni passed all its tests with flying colors, with only the one hiccup when it got entangled in a low-hanging cable during a vacuum-only session.

Read more about how we test vacuum cleaners

[First reviewed August 2025]

3i G10+ robot vacuum: two-minute review

The 3i G10+ is a hybrid robot vacuum and mop that has a clever USP that makes it an especially good choice for small homes: the onboard dust cup is around twice the usual size, and compresses its contents to fit even more in. It's an ideal solution for those people who don't have space for a massive self-empty base station taking up their living room, but equally don't want to have to manually empty dust after every clean.

While it lacks the fancier features found on today's best robot vacuums, it's certainly one of the best budget robovacs I've tested, punching above its affordable price tag. There are several features I wouldn't expect in this price bracket, including an extending side brush and a mop pad that kicks out to get closer to the edges of rooms and can lift when traversing carpet. The result is very thorough vacuuming, strong edge cleaning, and solid mopping too.

It's not perfect. I found the 3i G10+ struggled to mount the charger when the dock was placed on a carpet, and the object avoidance is a little hit-and-miss. It's also not the best choice if mopping is your priority – the small onboard tank needs refilling regularly, and the fact the mop pads can't be removed from their base for machine-washing is annoying.

However, given the overall balance of performance to price, I think the 3i G10+ is a very solid recommendation. For me, the pros vastly outweigh the cons. Read on for my full 3i G10+ robot vacuum review.

3i G10+ robot vacuum review: price & availability

• List price: $449 (about £560 / AU$1,110)
• Available: US only
• Launched: January 2025 (on sale June 2025)

The 3i G10+ was unveiled at CES 2025 and then went on sale in June that year. It's only available in the US, and any way you slice it, the 3i G10+ is very well-priced for what you're getting. At full list price, it's $449, but at time of writing it was reduced to $399 (about £560 / AU$1,110). Both fall into the TechRadar budget price bracket for robovacs. Sure, you can go cheaper with some robot vacuum models, but to get one with mopping capability, this is about as low as you can go.

There's also the innovative dust compression, enabling you to avoid having to make space for a bulky dock. There are only a few other bots that offer this. One example is the Narwal Freo X Plus, which costs about the same as the G10+ (going for $399.99 / AU$899), but has some unwelcome quirks that the 3i doesn't suffer from. There's also the Roomba 205 DustCompactor Combo for $469.99 / £399, but the least said about that bot and its performance, the better.

The Shark Matrix Plus 2-in-1 goes for the same price when discounted, but has a higher MSRP of $699.99 / £449.99. It does have a bigger auto-self emptying base station with a 1.9 quart (about 1.8L) dustbin capacity, so requires a bit more space. It's also a bit older of a model, which at the time of review had a glitchy app and just average mopping. With the 3i G10+, you get 1 still very good 1L dustbin without having to set aside the space for such a big base station. And that's Shark's most reasonably priced mopping robot vacuum.

The Ecovacs Deebot N10 Plus might be the robot vacuum that the 3i G10+ reminds me of most. The Ecovacs has a higher MSRP of $649.99 / £499 / AU$999, but can be regularly found at the same price as the 3i G10+. It also has a small auto-self emptying base station. It, however, has an annoying quirk in that you have to physically remove the mop pad for use on carpet. And in that regard, the 3i G10+ is easier to use, as I'll get into below.

While the 3i G10+ isn't quite perfect, it sets a high bar for anything around $400.

• Value for money score: 4.5 out of 5

3i G10+ robot vacuum specs

3i G10+ robot vacuum review: design

• Large capacity onboard dustbin with compression
• Water tank is small but easy to access
• Extendable side brush and mop pad

The 3i G10+ might have the typical disc shape, but it doesn't look like most robot vacuums. Instead of the little raised navigation tower poking out of the top, this robot vacuum has a much larger, almost egg-shaped raised portion that not only contains sensors (the typical reason for these raised areas), but also the power button and, most importantly, the dustbin.

That dustbin is unusual in that not only does it have a supersized 1L capacity – most robot vacuums have less than half of that – but also an internal mechanism to compress all the dirt and debris coming in. That means that even though there's no auto-empty dustbin in the dock (instead, it's a compact, charge-only setup), you won't need to empty the onboard bin too regularly. Because the dust is potentially sitting around in the bot for a while, 3i has built in a UV light to keep things hygienic.

Getting back to the exterior of the 3i G10+. The water tank hugs the back of the unit and pulls off and snaps back into place via magnets, making it very easy to access and fill up. It has a small capacity – just 247ml – so it needs regular refilling.

(If you're using the mop on a regular basis then a bigger, more comprehensive clean station that can wash and dry the mop pads for you, as well as refilling the tanks, is a big effort-saver. Although you will of course have to make space for it.)

Underneath, there's a single side brush, a brush roll, and a stationary mop pad. However, there's more going on here as well. That side brush can extend out to reach further into the edges of rooms. The mop pad is just a stationary one, but the 3i G10+ can move the whole pad out a bit for more thorough edge mopping too. Those kinds of features are unusual for robot vacuums at this price point.

• Design score: 4.5 out of 5

3i G10+ robot vacuum review: performance

• Very good vacuuming and solid mopping, including edges
• Mop pad can't be removed from mount for machine-washing
• Quiet operation

While I usually prefer robot vacuums to have two side brushes, one on each side, the single brush and brush roll on the 3i G10+ proved more than up to the job of vacuuming whatever dirt, dust, and debris was lying around. It helps that the side brush actually extends out, sweeping anything away from the edge of the room and into the bot's suction path.

I used this robot vacuum on hardwood flooring, tile and carpet, and found it did a very solid job on all three. I also ran TechRadar's usual suction tests, sprinkling flour, sugar, and breakfast cereal on the floor and seeing how efficiently it cleared it. It took two passes to get everything, but once those were complete, you couldn't tell that anything had been spilled.

The mopping performance is pretty good as well, though not quite up to the standard of the vacuuming. The mop pad is the flat kind that gets dragged across the floor. It's great for surface-level cleaning but not tough enough for anything more challenging – I find dual, spinning mop pads offer more thorough clean.

In general use I found the coverage was good, and not streaky the way that this kind of mop pad can be. Because the mop pad can kick out to the side, the si G10+ is great at mopping into the edges of rooms. Also a plus is the fact that it will lift the mop pad up a few millimeters when going onto carpet to keep said carpet from getting wet – you don't always get that kind of functionality with a bot at this price point.

I ran some mopping tests that involved spilling soda and coffee on the floor and seeing how well the bot handled it. While it cleared the area effectively, my whole floor ended up smelling like coffee and soda afterwards. I'd recommend tackling spillages manually, to avoid this.

Because the base station doesn't have mop-cleaning capabilities, you'll also need to stay on top of washing the mop pads fairly regularly. Annoyingly (and unusually), while the mop and its mount can be removed from the bot, the cloth part doesn't come off its mount. That means you can't throw it in a washing machine, but instead have to hand-wash. I was annoyed to find the mop cloth-and-mount took nearly a day to dry, though it went a lot quicker when I put it in the sun.

The object avoidance is unfortunately not as good as advertised. While the G10+ did manage to register and avoid some objects, I also watched it push around a bag clip, bump into a dog water bowl, and climb up on the base of a chair and end up stuck. Overall, the object avoidance is rather hit and miss. I've found that to be par for the course with more budget and mid-range models, though.

On the bright side, operation is relatively quiet, registering at 52dB during regular use and 56dB when extending side brush. I've used some robot vacuums that get much, much louder and it can be distracting. Having one with a quiet drone, especially when I'm trying to work or do something when it's running, is appreciated.

The battery lasts a very decent 240 minutes – enough time to clean the full 700 square feet area it could reach without having to recharge.

Lastly, though it vacuums just fine on carpet, it can't seem to get back onto its base station when on carpet (medium pile for reference). It's as if the 3i G10+ can't catch onto the ramp attached to the base station to reach the charging contacts. It's a bit strange and one of the drawbacks of not having a bigger mat for the robot vacuum to go up on.

• Performance score: 4 out of 5

3i G10+ robot vacuum review: app

• Straightforward to set up and use including maps
• Has an AI Auto Mode to make things even easier
• Comes with pet camera and manual controls

I typically don't find a lot of issues with robot vacuum apps these days, and it's no different with the 3i app. It's intuitive to get started, from pairing to going through the various instructions to complete the setup process. Other than waiting for the app and vacuum to pair and connect to WiFi, it's also a straightforward process.

Part of setting it up is also mapping your space and, though the 3i G10+ takes its time, it does a thorough job and even guesses the names of different rooms (not always correctly). Editing the map does take a tiny bit of menu diving, but once you get the screen to make changes, it works well and is happy to let you partition certain spaces more than once – an issue I've actually had with some models.

As far as using the app, the 3i G10+ home page is straightforward with the map top center, important info at the top, menu diving icons top right, as well as the more specific cleaning functions on the lower half of the screen.

There's a lot of flexibility for adjusting how you want your space cleaned including whether you want to vacuum or vacuum and mop, along with suction power, water level, cycles, and more. But if you don't want to manually set the cleaning mode, 'AI Auto' will decide the various settings for you. I found it useful for spaces with different types of surfaces – for example, it'll adjust automatically when going from hardwood to carpet.

Beyond the meat and potatoes of the app, there are more useful features including Alexa and Google Voice Assistant support. You can even take over the camera and motor functions to check on a pet when you're not home.

• App score: 4.5 out of 5

Should you buy the 3i G10+ robot vacuum?

Buy it if...

Don't buy it if...

3i G10+ robot vacuum review: alternatives to consider

How I tested the 3i G10+ robot vacuum

To test the 3i G10+, I used it for a couple weeks, testing its vacuum and mop capabilities not only on hardwood floors, but also when going from hardwood floors to carpets (with some intentionally dropped objects here and there). I also tested its cleaning abilities with spillages of various types. I went through every nook and cranny of the app, looking at things like mapping and ease of use.

Read more about how we test robot vacuum cleaners

• First reviewed July 2025

iRobot Roomba 205 DustCompactor Combo: two-minute review

The Roomba 205 DustCompactor Combo feels like a missed opportunity given iRobot's recent restructuring and slew of new upgraded floor bots. The unique dust compactor is an intriguing concept – the bot houses an oversized internal dustbin with a motorized paddle that compresses all debris, theoretically negating the need for a large self-emptying dustbin dock. However, in practice it's messy to deal with and ultimately impractical, especially if you have shedding pets in the home.

Granted, the 205's suction performance is acceptable enough – although not up there with what you'd get with one of the best robot vacuum on the market – its halfhearted mopping system is extremely outdated and when the mop pad is fitted, it prevents the robot from traversing rugs or carpet. The 205 is also a much noisier machine than any other robot vacuum cleaner I've thus far reviewed. And while its newly-adopted LiDAR navigation is clearly better than the company's previous V-SLAM camera-based incarnations, it also feels erratic and unsophisticated, at least when compared to other LiDAR systems.

I used to be a massive fan of iRobot products but I'm absolutely dumbfounded by how poorly its latest machines have behaved. The Roomba Combo 10 Max was mostly disappointing but I actually think that the Roomba 205 is worse. It's just so frustrating to use and requires too much intervention from the user. I wish I could recommend this model but it just doesn't perform as you'd expect for a robot vacuum cleaner in the year 2025.

Roomba 205 DustCompactor review: price & availability

• List price: $469.99 / £399
• Launch date: March 2025
• Availability: Widely, including US and UK

With a list price of $469.99 in the US and £399 in the UK, the iRobot Roomba 205 DustCompactor Combo sits towards the top of the mid-range price bracket. Note that at time of writing, there's a $170 discount in the US.

Given all the peculiarities with this machine, if you don't mind making a little more space for a simple, small-bodied, self-emptying dock in your home, I would recommend doing some research on other models in the same price range. For instance, the Roborock Q5 Pro+ has a list price of $699.99 / £499.99 but can often be picked up for a similar price to the Roomba 205, and offers far more reliable performance.

Alternatively, if a small dock is the priority and you're not dealing with huge amounts of dust, it could be worth considering a model without dust compression. There are a number of cheaper options in this category, for example, the Dreame D9 Max Gen 2, which costs $299.99 / £219 at list price and impressed our reviewer.

If you're hellbent on going for the Roomba 205 DustCompactor Combo simply because of its unique bin system, by all means go ahead. But make sure your home is on the smaller side and with no rugs or carpet to block its path when the mopping pad is attached.

• Value for money score: 2.5 out of 5

Roomba 205 DustCompactor specs

Roomba 205 DustCompactor review: design

• Hybrid robot vacuum and mop with internal dust compression
• Compact, charge-only dock that refuses to stay put
• Only one rubber roller rather than iRobot's usual double rollers

Given the financial troubles that have besieged US brand iRobot during the past few years, it's good to see that the company has since restructured itself and developed a new range of robot vacs and mops entirely from the ground up. However, in design terms, I feel that the Roomba 205 is a pale shadow of its forebears like the Roomba ‘i' and ‘j' series, and especially the excellent S9+ (for my money, by far and away the best robot the company ever made).

The USP for the Roomba 205 DustCompactor Combo I'm reviewing here is that it's equipped with an extra large onboard dustbin (about 1.5 liters) with an internal motorized plastic paddle that swings back and forth to compact all detritus towards the back of the bin, thereby creating more space for subsequent cleaning sessions. That means it can offer a large dust capacity without the need to make space for a large auto-empty dock.

iRobot states a period of about 60 days before the bin needs emptying but I can see it filling up with dust and hair within a week or so if you have shedding pets in the home (this is not one to consider if you're on the hunt for the best robot vacuum for pet hair). On the plus side, you will never need to buy any extra dust bags.

From an aesthetics point of view, the 205's dock is almost invisible in a room so I can see the appeal of a back-to-basics approach like this. However, from my tests I've found that emptying the bin myself has been quite messy and not especially pleasant (I'll get into this more in the Performance section).

Personally, I'm not sure such an innovation is necessary, either. The are models on the market that have petite, unobtrusive self-empty docks equipped with large dustbin bags – I'd happily sacrifice those few more inches of floorspace for the relative ease of this setup.

Another drawback of iRobot's solution is that the 205's titchy dock is so light in weight that it shifts around when the robot is docking and this often causes the robot's contacts to fail to connect with the dock. It's even worse when the dock is placed on carpet – I tried this the first time I used the 205 and it failed to dock at all; the charging base simply refused to stay put.

Let's take a look at the unit itself. The Roomba 205 is roughly the same width as most robot vacs but a bit taller, which means it might not reach under some low furnishings.

Flip it over and you'll notice that the 205 has only one rubber roller brush, which is a bit strange considering that iRobot invented the twin roller brush that so many high-end robot vacs use these days. I'm not sure why, but this roller brush tends to clatter across my hard wooden floors and jitter over one of my Persian-style rugs as if it were on a corrugated surface, which is distinctly annoying.

iRobot doesn't list any of its robots' suction power in Pascals so it's difficult to tell how powerful it is in comparison to other bots on the market. But, hey, you can take it from iRobot that it's ‘70x the suction of the older 600 series', whatever that means. Incidentally, you get three levels of suction power with this unit – Light, Normal and Deep.

The 205 also comes with the obligatory side spinning brush for flinging wayward debris along edges into the path of its suction. However, unlike many newer (and perhaps pricier) robovacs with side brushes that automatically articulate outwards when traveling along edges and in corners, the 205's side brush is firmly fixed in position.

The 205 DustCompactor's mopping system comprises a simple mop pad with its own 250ml water container above it. The user takes charge of filling the tank and washing the pad – the dock won't help with this, as is the case with some bots.

For mopping, the robot will simply drag the pad along the floor. You can set the amount of water that's dispensed but just don't expect any miracles, especially when it comes to deep stain removal. This kind of basic mopping setup isn't unusual setup for an affordable robovac, and for general day-to-day floor cleaning it does pass some muster.

Me? I've removed it entirely because I've since discovered that the robot refuses to cross rugs or carpet if the mopping pad is attached, but more on that in the Performance chapter.

Given that iRobot has refused for years to include LiDAR as part of its navigation system, even when every other manufacturer and their dogs have, it's refreshing to note that the company has at last seen fit to include LiDAR navigation on the 205.

However, while most bots have their LiDAR turret in a raised puck their top surface, to allow a full 360-degree view, here iRobot has hidden it behind a narrow slot just above the camera, thus restricting the view to around 180 degrees. Thankfully the camera provides backup, although it's not AI enabled like so many models these days.

Setup

The iRobot Roomba 205 DustCompactor Combo arrives in a small box containing the robot, its tiny charging dock, a clip-on side brush and rear mop mount. Initial setup was relatively easy but only once I'd discovered that the old iRobot Home app isn't compatible with this model, so I downloaded the newly upgraded Roomba Home app instead.

All went according to plan until it connected to my Wi-Fi and then proceeded to automatically download the latest firmware update. That's normal, I thought, but then a warning popped up saying that ‘Roomba is having some trouble updating and to make sure Rosie (Rosie?) is docked with a good Wi-Fi connection'.

‘Rosie' was docked and the internet was fine so I waited a while, then got impatient and restarted it. It sprang to life again and gave the impression it was ready to work, so I commanded it to and it proceeded to spin round on its dock for about a minute, entangling itself on its own power cable.

I took it off the dock and placed it on the floor, whereby it started spinning again. At this point I thought I'd been handed a dud so I took it to a different area downstairs, plugged it in and left it to fully charge. Lo and behold, when I launched the app again, it showed the ‘start mapping' screen which suggests it was updating all along, only without any icons or progress bars to tell me so.

Anyway, I tapped the mapping button and it proceeded to map my entire floor plan efficiently and – because the roller brush was turned off – without getting caught on any rugs. The upshot of all this is that if you decide to buy a Roomba 205, just be patient and let it fully charge and complete its first firmware update, even if it says it's having trouble connecting, before trying to request any cleaning.

• Design score: 2 out of 5

iRobot Roomba 205 DustCompactor Combo review: performance

• Decent vacuuming but emptying is messy
• Woeful mopping and won't ride over carpet with mop attached
• Below average obstacle avoidance

After a confusing setup, the 205 started behaving as expected and mapped my open-planned, four-room downstairs area surprisingly quickly, even if it did create nine different rooms that I had to merge in the app. This is common with most robot vacs when mapping open-plan environments though it has to said that nine divisions is a record for me.

Also, for a LiDAR-equipped robot, the detail of its mapping in the app is lamentable and way behind the likes of Roborock, which displays every nook and cranny instead of just a basic overview of one's floor plan.

Since it's been running, the 205's LiDAR-based navigation has been mostly okay though I have had one or two issues where it has either got stuck while trying to cross a loose rug – a common issue with some models – or suddenly stopped dead next to a kitchen cupboard for no apparent reason. With most other robot vacuum I've tested, I've learned their working methods to some degree, but this model's navigation process remains largely a mystery to me.

Obstacle avoidance

As mentioned above, the Roomba 205's LiDAR is tucked under a slotted housing so it doesn't appear to give the robot a full 360–degree view. There is a camera on the front, however, and this seems to help with its obstacle avoidance. In my test I placed the usual assortment of dog toys in front of it, plus the obligatory phone cable. It carefully avoided the first two large toys but simply snow-ploughed all the others. As expected, it also rode over the phone cable and then dragged it along behind it for a while.

In the pantheon of obstacle avoidance, the Roomba 205 isn't the worst I've seen but it's not that great either, and clearly not a machine to have running around if you have (for example) a non-house trained puppy in the home.

Vacuum performance

Let's take a look at the vacuuming performance. There are few niggles to note up top. Firstly, as noted above, the Roomba 205's rubber roller made quite a racket as it trundled across my wooden hard floors. But perhaps worse than that was the monotonous high-pitched sound of the suction motor, which was clearly audible from meters away. I measured the sound waves at an average 78dB from five feet away and at an annoyingly high-pitched frequency of 660Hz.

Aside from the noise issues, the side sweeping brush fell off twice during my tests, despite being firmly connected before setting off.

To get an objective ideal of this bot's suction powers, I ran TechRadar's standard vacuuming tests. This involved sprinkling debris of different sizes – specifically, flour, muesli and rice – on the floor and assessing how efficiently the robot picked it up. Here, the Roomba 205 was pretty decent, though it did leave traces of flour in the cracks.

However, when I later went to the dock to grab a photo of the robot returning after a standard cleaning session, I noticed a small scattering of rice and muesli that it had deposited in front of the dock from its previous cleaning test, which suggests it never fully sucked all the rice and muesli into its bin and instead simply held some of it in its brush housing. This anomaly hasn't repeated itself on subsequent tests, so hopefully it was just a glitch.

I then proceeded to run the same tests on carpet. This proved to be quite a challenge, because although the app has a menu item called ‘furniture and zones' under the heading ‘routines', nothing happens when you tap it. What it should do is provide an editable square that can be expanded or contracted over a specific cleaning area.

In the end I opted to select ‘spot clean' which was equally tricky because the map for each room is very basic and cannot be expanded so it takes trial and error to find the exact spot that relates to a specific area. Anyway, I'm pleased – indeed surprised – to report that the 205 collected all evidence of rice and muesli though, like other robots I've tested using this method, it did leave some flour marks behind.

Dust compactor performance

Moving on to the unusual bin system. While the Compactor's paddle does indeed push all detritus to the outer edges of the bin, I found that emptying it has been quite messy and not especially pleasant.

Firstly, the lid of the robot's bin doesn't open anywhere near wide enough, so I couldn't find a practical solution for emptying its dusty and hairy contents into my average-sized Brabantia kitchen bin without some of the contents spilling all over the floor. Worse, my model has a sticky bin latch so I've had to be doubly careful when trying to open the bin lest its contents end up all over the floor and my clothing.

iRobot states ‘up to 60 days' of cleaning before the dustbin requires emptying but my test model required emptying after just eight days of vacuuming. I should add that I have two Labradors and two cats that regularly shed hair and my home is also surrounded by woodland and farm fields so it's dustier than most urban environments.

Mopping performance

Onto the mopping. First things first: if you have any rugs or carpet dividing your hard floors, this may be the last hybrid robot you want, because it will not traverse any rug or carpet when the mopping pad is fitted. I first thought that it was the small protruding lip at the bottom of the mop mount that was catching on my rugs but it transpires that the unit has a carpet sensor on board that, in my case, refused to allow the robot to proceed the moment the front half of the robot touched the rug.

It did do its darnedest to head over the rug en route to the hard floor kitchen, but after 10 minutes of trying at every conceivable angle, it eventually announced that it was unable to continue cleaning before returning to the dock, having not performed the assigned mopping task. I can't believe that I'm the only household in the world with rugs scattered over hard flooring but what I do know is that, for me, this model has proved to be as useless as a chocolate teapot when it comes to complete household cleaning without any human interjection.

Anyway, having failed to ride over my rugs, I resorted to carrying the bot to the kitchen so it could perform a spot clean on a spill of milk. I had to make several attempts at this because, as I had discovered for the carpet vacuuming test, the app's map is so unrefined that it's almost impossible to select a specific cleaning spot. Anyway, I eventually spilled some milk in its path and it duly soaked up the milk leaving no signs of it. But then I expected that – much like I'd expect a single swipe with a paper towel to do the same thing.

In its favor, it did also mop the rest of the area quite effectively, leaving only a damp swathe that dried up quickly. I then had to pick the robot up again and place it near its dock so it could get back to charging.

It's not unusual for a budget robot vacuum to have this kind of simple mopping setup (not including the refusal to pass over carpet – that is very unusual). More advanced bots will have pads that vibrate, rotate or press down to dislodge dirt, but you might need to pay a little more for this kind of setup.

• Performance score: 2.5 out of 5

iRobot Roomba 205 DustCompactor Combo review: app

• New app for this current range of Roombas
• Very basic with incomplete menus and unrefined maps
• Frustrating and slow to use

The new iRobot Home app was recently launched as a replacement for the older app that's now been renamed iRobot (Classic). Having used it extensively during the past week I can report that there's a lot of work to be done on making this app more user-friendly. Firstly, it's slow to send commands to the Roomba 205 and some user inputs cause a spinning cog wheel to pop up on screen while it sends its signal.

Perhaps I'm too used to Roborock's fully comprehensive app but I found the iRobot Home app to be too simplistic and, at times, very irritating to use. For instance, the floor plan maps it creates are so basic it's difficult to tell where one room ends and another begins. Also, its Spot Clean function is almost impossible to use since you can't expand the map in this menu and the little spot cleaning square it provides can't be placed exactly where you need it.

Although the app allows you to create routines and standard cleaning schedules, and make adjustments to both vacuum power and water amount for mopping, you can't create any specific cleaning zones, like in front of the dog bed or a specific rug. Bizarrely, there is an option to select ‘Furniture and Zones' but when you click on this function, nothing at all happens. I'm sure iRobot will be adding improvements to this app going forward but, as it stands, it's woefully inadequate.

• App score: 1.5 out of 5

Should you buy the iRobot Roomba 205 DustCompactor Combo?

Buy it if...

Don't buy it if...

How I tested the iRobot Roomba 205 DustCompactor Combo

As is the way with all TechRadar tests, we live with the product for as long as possible, observing its day-to-day functions and generally appraising it from afar. We then get down to some serious testing while making copious notes on the product's performance.

In the case of the iRobot Roomba 205 DustCompactor Combo, I had it installed at home for just over a week while I observed its navigation behaviour. I then proceeded with my vacuum, mopping tests on both hard floor and carpet using a variety of household ingredients in order to replicate a real-world scenario. I then tested its obstacle avoidance system using some dog toys and the obligatory phone cable.

I should add that this particular review took me a lot longer to complete than usual because the machine in question failed to behave as I expected. Hence, I had to spend extra time repeating tests to make sure its previous hiccups weren't just one-off glitches.

Read more about how we test robot vacuum cleaners

• First reviewed July 2025