How to choose a robot vacuum in 2026: the questions that actually matter

Start with a floor survey, not a spec sheet

Robot vacuum listings compete on the numbers that are easiest to print: suction in pascals, battery minutes, the number of things the dock does. None of those tell you whether a particular robot will cope with your home. The more useful first step is a short survey of the rooms you want cleaned, done with a notepad and a tape measure.

Vacuum Wars, which publishes lab-based robot vacuum rankings, opens its 2026 buying guide with the same idea. Before comparing models, it suggests noting how much carpet you have, whether pet hair or long human hair is a daily problem, how cluttered the floors usually are, which rugs must never get wet, how much clearance your furniture leaves and whether any thresholds sit between rooms. Its summary line is to choose for the conditions in your home, and it says plainly that the most expensive robot is not automatically the best fit.

Write your answers down in one line each. By the end of this guide, each line will either rule models out or tell you which feature to pay for.

Measure the three numbers that rule robots out

Height comes first. Vacuum Wars puts a standard robot at about 102 mm (4 in) tall and the lowest models it has seen at about 80 mm. Narwal's own buying guide adds that a laser turret on top of a LiDAR robot adds 1 to 2 cm, and that typical LiDAR models need roughly 10 to 12 cm under furniture. Measure under your lowest sofa, bed and cabinet run at a few points, because sagging cushions and uneven floors change the answer. If the dustiest floor in your home sits under a low sofa, a model such as the Roborock Qrevo Slim, listed at 8.2 cm tall, is built for exactly that (Roborock Qrevo Slim review), and the Roborock Saros 10R takes a camera-led, low-body approach to the same problem (Roborock Saros 10R review).

Thresholds come second. Vacuum Wars describes the traditional climbing limit as about 20 mm (0.79 in), and Narwal's guide gives a similar 2 cm figure for most robots. Some newer models lift their chassis or use small legs to climb higher, but Vacuum Wars warns that the biggest advertised climbing figure often refers to a stepped, two-tier threshold rather than a single lip. If you currently lift a robot over a doorway, the Dreame L50 Ultra's ProLeap system is one design aimed at that specific problem (Dreame L50 Ultra review).

Width and access come third. Chair legs, a crowded desk or a dining set can leave a full-size robot cleaning around the edges of a room rather than under it. Compact designs exist for this; the SwitchBot K10+ Pro is 248 mm across (SwitchBot K10+ Pro review), though small robots usually give up mopping and dock features to get there.

Decide how much of the routine you want to hand over

The dock decides how much work remains for you, and it accounts for much of the price gap between models. Vacuum Wars groups docks into three levels. A charging-only base is the cheapest and smallest, but you empty the bin, refill the water and wash the mop yourself. An auto-empty dock pulls debris out of the robot into a bag or bin. A full-service mop dock also washes and dries the mop pads and refills the robot's water, at the cost of a bigger footprint and its own tanks and trays to clean.

Be honest about which chore actually stops you running the robot. If it is emptying a small bin, an auto-empty dock fixes that without the plumbing-like upkeep of a wash station. If you mainly have carpet, a mop dock is mostly expense. If mopping a busy kitchen every day is the goal, the full station is where the money should go. Our companion guide on dock types covers this choice in more detail.

Hair, carpet and the brush question

On hard floors, most current robots pick up loose debris well enough that the differences are small. Carpet and hair are where they separate. Vacuum Wars reports that bristled brushes agitate carpet better than all-rubber ones, and that split or tapered rollers let hair travel off the brush instead of wrapping around it. It also treats an onboard bin under 300 mL as small for a pet home.

Vacuum Wars is also clear about limits: it describes a robot as a frequent maintenance cleaner rather than a replacement for periodic deep cleaning with an upright or canister vacuum, because embedded grit in carpet still favours a full-size machine. Plan on keeping your regular vacuum for that job. If carpet is most of your floor and you do not need mopping at all, a vacuum-only robot such as the Roborock Q5 Max+, with dual rollers and a 770 mL onboard bin, puts the whole budget into pickup (Roborock Q5 Max+ review).

Obstacle avoidance and the camera trade-off

Navigation and obstacle avoidance are different jobs. Navigation, usually LiDAR or camera-based mapping, lets the robot clean in tidy rows and return to its dock; Vacuum Wars notes that even inexpensive models now do this reasonably well. Obstacle avoidance is about the cable, sock or pet accident sitting below the robot's laser line.

Vacuum Wars publishes average results from a 24-object course, and the spread is wide. Robots relying on a single-line laser averaged 3.6 objects avoided, structured light alone averaged 7.1, and systems pairing sensors with one or two cameras averaged 18.5 and 20.1. That gap is the strongest argument for a camera in a home with pets, young children or cables on the floor. It is also a privacy decision, and Vacuum Wars treats preferring a camera-free robot as a legitimate trade-off. Whichever you choose, no system is foolproof, and a quick pick-up before each run remains the most reliable way to avoid a chewed charging cable.

Set a budget by tier and ignore the suction race

Vacuum Wars sorts the US market into price bands. Under US$300 brings the biggest compromises in navigation, mopping, obstacle avoidance or automation. From $300 to $600 you get reliable basic cleaning, and some models add smart navigation, a simple mop or an auto-empty dock. It calls $600 to $1,000 often the value sweet spot, with better sensors, more capable mopping and more automated docks, while bands above $1,300 add flagship mobility and the newest dock features, with diminishing returns beyond $1,600 for typical homes. Canadian shelf prices are higher in dollar terms, so use the bands to understand what each step buys rather than as a Canadian price list.

Treat suction as a minimum, not a ranking. Narwal, a manufacturer with every reason to sell large numbers, writes in its own guide that a 20,000 Pa robot does not clean twice as well as a 10,000 Pa one, because airflow, brush design and floor contact matter as much. A record-setting figure is the least reliable reason to pay more.

A five-step shortlist

First, write down your clearance, threshold and access measurements and strike out any model that fails them. Second, choose the dock level that removes the chore you actually skip. Third, if hair or carpet matter, require a split, tapered or bristled anti-tangle brush and look for independent hair-pickup results rather than marketing percentages. Fourth, decide whether you want a camera, and if you do not, plan to tidy before every run. Fifth, price the full package, including bags, pads and brushes; our running-costs guide covers that arithmetic.

At that point most homes are down to a handful of candidates. Buy from a retailer with a return window long enough to run the robot on your own floors, and pick the cheapest model that clears every line on your survey. A budget design such as the Dreame D9 Max Gen 2, with no big station at all, can be the right answer if you will genuinely empty its bin (Dreame D9 Max Gen 2 review).

From a tech standpoint

The best robot vacuum is the one that physically fits your home and removes the chore you keep putting off. Measure first, choose the dock second and let suction figures come last. A model that clears your sofa and climbs your doorway will do more for your floors than a higher number on a box.