Robotic pool cleaners in 2026: cordless vs corded, coverage, filtration and runtime compared

This guide explains the specs that separate one robotic pool cleaner from another: power source, what surfaces it cleans, runtime, filter fineness, pool-size ratings, weight, retrieval and navigation. It draws on all 33 pool robots in our database, shows which figures makers publish and which they leave out, and ends with a shortlist method by pool type and budget.

Last checked 1–9 October 2026. Spec-based analysis from maker-published figures; we have not tested these robots hands-on.

The short answer

Of the 33 pool robots we track, 29 are cordless and four plug into the mains. Cordless now dominates the market, but it is not automatically better. A corded robot never runs flat mid-cycle. A cordless one needs charging, has a fixed runtime and can be harder to fish out.

Five questions narrow the field fast:

1. Above-ground or in-ground? Seven robots in our data are rated for in-ground pools only. Two are rated for above-ground pools only.
2. Floor only, or walls and waterline too? Three robots clean the floor only. Twenty-eight list the waterline.
3. Do leaves float on your pool for hours? Only eight robots list surface skimming, and all of them cost $999 or more at their lowest maker price.
4. How fine is your debris? Sand and dust need a fine filter. Leaves need a big basket. Few robots do both well from one filter.
5. Who will lift it out? Robots in our data weigh up to 13.2 kg dry, and more with water inside.

Browse all records in the robot pool cleaners category, filter them in the finder, or put two side by side in compare.

All 33 robots at a glance

Coverage key: F = floor, W = walls, WL = waterline, S = surface skimming. Runtime is in minutes, as the maker splits it by mode. "n/p" means not published. Prices are the lowest maker-store price in our record, with list price in brackets where a sale was shown. They change often.

Our database stores sizes in square metres and metres, weights in kilograms and times in minutes. Each robot page shows the maker's original figure and source link.

Cordless vs corded: what you actually trade

Corded robots run from a power supply that sits on the deck. The robot gets low-voltage DC down a floating cable. Maytronics' manual for the Dolphin Sigma lists an output of "<30VDC" (Maytronics manual 8180035). Zodiac's manual for the Voyager RE 4600 iQ states 30 V DC to the cleaner and a 150 W maximum load (Zodiac Voyager RE manual).

The cable is the main drawback. It can twist, and it limits reach. The Dolphin E10 page lists a 12.0 m cable and no anti-tangle swivel (Maytronics E10 page). Zodiac's cable is 15 m or 18 m with a tangle-reducing swivel. Zodiac is blunt about removal: "Never lift the cleaner out of the pool by the floating cable."

Corded cycle times in our data run from 90 to 150 minutes. These are programme lengths, not battery limits. You can start another cycle straight away.

Cordless robots remove the cable and the deck-side box. But they add charging, a battery that ages, and a runtime ceiling. The charger still has rules. The Dolphin Liberty 400 manual says to place it at least 3.5 m from the pool edge, use a circuit with a 30 mA residual-current device, and not to use an extension cord (Maytronics Liberty manual). The corded Sigma and Zodiac manuals set the same 3.5 m distance for their power supplies. So "cordless" does not mean "no socket near the pool". It means the socket can be in the garage.

All three manuals we read say no one should swim while the robot is in the pool.

Coverage: floor, walls, waterline and surface

Coverage is the clearest dividing line in this category.

• Floor only (3 of 33): the Aiper Scuba E1, Aiper Seagull SE and Dolphin E10. These suit small, flat above-ground pools. Maytronics says the E10 also reaches the angle where floor meets wall, but its spec sheet marks waterline scrubbing as "No".
• Walls (29 of 33) and waterline (28 of 33): most robots now climb. Wall climbing matters because algae and grime build up on vertical surfaces. Waterline scrubbing matters because oils and scum collect in that band.
• Surface skimming (8 of 33): the robot floats and collects leaves, insects and pollen before they sink. Beatbot and Aiper make all eight in our data.

Watch for gaps. The Ecovacs Ultramarine P1 and Nautilus CC Plus list floor and walls, but not the waterline. The Roborock RockAqua P1 page describes floor and waterline modes but does not state wall cleaning (Roborock RockAqua P1 page). The WYBOT S2 is the only record that lists steps as a separate surface.

Runtime and charging: read the mode, not the headline

Pool robot runtimes are easy to misread. Skimming is gentle work, so surface runtimes are long. The eight surface-capable robots quote 420 to 660 minutes in surface mode. Their floor runtimes, where published, are all 300 minutes. The 660-minute figure belongs to the discontinued AquaSense 2 Pro.

Across all 33 records, 32 publish some runtime. But only 12 split it by mode. Eleven give one figure with no mode at all. The rest name a single mode such as "eco" or "standard". Eco modes run longer because they clean less hard. The Aiper Scuba S1 is a good example: 180 minutes standard, 270 minutes eco.

For a fair comparison, use the longest non-surface figure. On that basis the median across 31 records is 180 minutes. The range is 90 minutes (Seagull SE and E10) to 300 minutes (nine Aiper and Beatbot models).

Charging is slow. The 25 records with a charge time range from 150 to 360 minutes, with a median of 240. The Liberty 400 takes "up to 6 hours" to charge for a 150-minute run. Its manual lists a 126 Wh lithium-ion battery. Maytronics adds that the charger only works between 6 °C and 35 °C. Few makers publish battery capacity at all. In our data, only the Liberty 400 (126 Wh) and the AquaSense 2 Pro (13,400 mAh) have it.

Some marketing claims are not runtimes. Aiper's Scuba V3 page promises "7 days" of carefree cleaning but gives no runtime figure (Aiper Scuba V3 page). Beatbot's AquaSense X quotes 3,875 sq ft of coverage "on a single charge" (Beatbot AquaSense X page). That is an area-per-charge claim, not a maximum pool size.

Filtration: what a micron rating does and does not tell you

A micron (µm) rating is the smallest particle the filter mesh is meant to hold. A lower number means finer filtering. Leaves and insects are caught by any filter. Fine sand, dust and pollen need a fine mesh. The trade-off is clogging. Fine mesh blinds faster and may need rinsing after every cycle.

Twenty-five of our 33 records publish a micron figure. The median is 150 µm. Ten robots list 180 µm, eight list 150 µm, and two each list 250, 70 and 3 µm. One lists 10 µm.

Those numbers are not like for like. Many robots use two filter stages. Our record stores one headline figure, and makers choose which stage to headline:

• WYBOT C2 Vision: the maker page lists 180 µm cartridges plus 10 µm ultra-fine filters (WYBOT C2 Vision page). Our record shows 180.
• Zodiac RE 4600 iQ: the maker page lists "dual-stage 150/60µ filtration" (Zodiac RE 4600 iQ page). Our record shows 150.
• Roborock RockAqua P1: the page lists 180 µm plus 70 µm dual-layer filtration. Our record shows 70.
• Aiper Scuba X1 Pro Max: our record shows 180 µm. A 3 µm MicroMesh filter is sold separately at $59.99, and was listed out of stock when we checked (Aiper accessory page).
• Aiper Scuba X1: the page gives only the 3 µm figure and no rating for the basket (Aiper Scuba X1 page).

So before you compare micron figures, check which stage each one describes and whether the fine stage comes in the box.

Basket size matters for leafy pools. Twenty-two records publish a filter capacity. They range from 2 L (AquaSense 2) to 6.5 L (both Degrii models), with a median of 3.7 L. The AquaSense X adds a docking station that rinses the filter and empties debris into a 22 L bin. The station weighs 41.9 lbs on its own.

One more caution comes from the manuals. Maytronics warns against running the robot after liquid flocculant or clarifier has been added, because these damage the filters.

Pool size ratings: area, length and shape

Makers rate pool size in two ways, and they do not convert neatly.

Twenty-seven records give a maximum area. Twelve give a maximum length. Only the Roborock gives neither. All five Maytronics robots give length only, from 8 m (E10) to 15.24 m (Sigma and EON 120d). Zodiac gives both: 12 × 6 m, which our record stores as 72 m² and 12 m.

Area ratings range from 72 m² to 464.52 m² (Degrii Zima Pro), with a median of 300 m². Treat them as ceilings. A long, narrow pool may exceed the length rating well before it reaches the area rating. Deep ends, steps, ledges and sun shelves also add surface the floor-area figure ignores. Two area figures in our data are unconfirmed: the Zima Lite figure comes from a retail listing, and the AquaSense X figure could not be matched to a maker spec line.

A useful rule: pick a robot rated at least one size above your pool if you want wall and waterline coverage in a single cycle.

Weight and retrieval: getting it out of the water

Twenty-five records publish a dry weight. The median is 9.5 kg. The range runs from 3.45 kg (Seagull SE) to 13.2 kg. Surface skimmers are the heaviest group. The six with a maker-stated weight range from 10.5 kg (Sora 70) to 13.2 kg (AquaSense 2 Ultra). No maker publishes the weight with water inside, which is what you actually lift.

Retrieval features solve that problem in different ways. Nineteen of 33 records describe one:

The details matter. Maytronics' Click-Up works by shaking a "Clicker" in the water for more than three seconds. The robot then climbs its nearest wall, and the feature stays available for two days after a cycle. Zodiac's Lift System lets you rotate the robot to face a chosen wall before it climbs, and the manual says to use it only with the cleaner in view. Roborock says its robot returns to the waterline when cleaning ends or the battery drops below 15%.

Navigation: ultrasonic, camera and trademarks

Navigation claims are the least comparable part of a pool robot spec sheet. Of the 27 records that describe navigation, most give a trademark or a path pattern rather than a sensor. Examples include WavePath, FlexiPath, SmartMap, 3D-Navi and "S-path and N-path".

Where a sensor is named, three types appear:

• Ultrasonic (sonar): named in 10 records, all from Beatbot, Degrii and Aiper. Ultrasonic sensors measure distance to walls and obstacles by sound, which travels well in water. Aiper says the X1's sensors slow the robot before obstacles.
• Camera or vision: named in five records: the Aiper Scuba V3 and V3 Ultra, Beatbot AquaSense X, WYBOT C2 Vision and Roborock RockAqua P1. Cameras are used to spot debris and steer towards it. WYBOT claims "20X Cleaning Efficiency", but a footnote limits this to its floor-only AI Vision mode.
• Time-of-flight (ToF): named in two records, the Aiper Scuba S3 and AquaSense 2 Ultra. ToF sensors time a light pulse to measure distance.

No record in our database names a gyroscope or inertial sensor. Many robots likely use one for wall turns and heading, but makers do not say so.

Sensor counts are marketing, not measures. Beatbot quotes 29 sensors on the AquaSense X and 27 on the P300 Ultra. Aiper lists 14 on the X1 without saying what they are. A count tells you nothing about what each sensor does.

App control also has a physical limit. Radio is strongly attenuated in water, which is why underwater networks use sound (Partan, Kurose and Levine, 2007). That helps explain why many robots only report status when they surface or park. Aiper sells a separate HydroComm device as an underwater communication relay for its X1 series (Aiper HydroComm page).

Above-ground vs in-ground, and what your water allows

Twenty-six records state which pools they suit. Eighteen include above-ground pools. Seven are in-ground only, including the Aiper Scuba X1, X1 Pro Max, Scuba V3 Ultra and Dolphin Sigma. The Scuba E1 and Dolphin E10 are rated for above-ground pools only. The discontinued Seagull SE adds in-ground pools with flat floors only.

Above-ground pools raise two issues: soft vinyl walls and getting the robot over a tall side. Light floor-only robots suit them. Heavy wall-climbers may not.

Check depth limits too. The Liberty 400 manual sets 0.3 m to 3 m. Beatbot says the AquaSense X works in water as shallow as 14 inches.

Water chemistry limits are published in manuals, not on product pages. Maytronics' Liberty and Sigma manuals set the same limits: chlorine 4 ppm maximum, pH 7.0–7.8, salt (NaCl) 5,000 ppm maximum and water at 6–35 °C. Sigma's manual adds that climbing may suffer below 15 °C. Zodiac's manual says not to run the RE 4600 iQ below 13 °C or above 35 °C. Beatbot rates the AquaSense 2 for salt water below 5,000 ppm. If you run a salt-chlorinator pool, check this before you buy.

Price tiers

Twenty-six records carry a US dollar price. Using the lowest maker price in each record:

Two records in the top tier are discontinued, so their list prices are history, not offers.

List prices are often fiction. Eighteen of the 26 US-priced records show a sale price below list. The median discount is 29%, and the range is 6% to 47%. The largest gap is the Aiper Scuba X1: $799.99 against a $1,499.99 list. Budget on the sale price, and expect it to move.

Warranty is a better long-term signal. Twenty-eight records publish one. Fourteen offer 24 months, nine offer 36 months and five offer 12 months.

How to choose by use case

• Small above-ground pool, tight budget: a floor-only robot is enough. Check that it is rated for above-ground pools and that you can lift it over the side.
• Typical in-ground family pool: a cordless floor, wall and waterline robot in the $500–$999 tier covers most needs. Compare the longest non-surface runtime against your pool's size rating.
• Pool under trees: weigh surface skimming against a cheaper robot plus manual skimming. Surface capability starts at $999 in our data and the median is far higher.
• Sandy or dusty site: look for a published fine stage (10 µm or below) and confirm it ships in the box.
• Large or long pool: check length ratings, not just area. The X1 Pro Max states 30.48 m, the longest in our data.
• You never want to charge a battery: a corded robot such as the Nautilus CC Plus or Zodiac RE 4600 iQ runs cycle after cycle. Plan for cable handling.

For curated picks, see our best robot pool cleaners list and the robot pool cleaner buying guide.

What the numbers say (Analysis)

• Cordless has won the catalogue, not the argument. 29 of 33 records are cordless. The four corded robots run 90–150-minute programmed cycles that a battery cannot cut short, while cordless robots need a median 240 minutes to recharge (25 records).
• Surface skimming is a premium feature. 8 of 33 records list it. Their median lowest US price is $2,099.50, against $674.50 for the 18 US-priced robots without it. The cheapest skimmer in our data is the Beatbot Sora 70 at $999.
• Most runtime figures cannot be compared directly. 32 of 33 records publish a runtime, but only 12 split it by mode and 11 give a single figure with no mode. The median best non-surface runtime is 180 minutes (31 records); surface modes reach 420–660 minutes.
• Micron ratings are reported unevenly. 25 of 33 records publish one (median 150 µm). At least five robots use two filter stages, and makers headline the coarse stage on some and the fine stage on others.
• Weight and battery data are thin. 25 of 33 publish a dry weight (median 9.5 kg, range 3.45–13.2 kg). None publishes a wet weight. Only 2 of 33 give a battery capacity, and only 9 of 33 give an IP rating.
• Discounts are routine. 18 of 26 US-priced records show a sale below list, with a median discount of 29%.

Frequently asked questions

Is a cordless pool robot better than a corded one?

Not always. Cordless removes the cable and the deck-side power box. Corded robots never run flat and can run back-to-back cycles. In our data, corded cycles last 90–150 minutes, while the cordless median best floor runtime is higher, but cordless robots then need hours to recharge.

What micron rating do I need?

For leaves and insects, any rating in our data will do. For fine sand, dust or pollen, look for a fine stage of 10 µm or below, such as the 10 µm WYBOT C2 or the 3 µm Aiper S3 and X1. Finer filters clog faster, so expect to rinse them more often.

Do I still need a skimmer if my robot cleans the waterline?

Waterline cleaning scrubs the band on the wall. It does not collect floating leaves. Only surface-skimming robots do that, and only 8 of our 33 records offer it.

Can I leave a robot in the pool while people swim?

No. The Maytronics and Zodiac manuals we read all say to remove the robot before anyone enters the pool.

Will a robot work in a saltwater pool?

Many will, within limits. Maytronics' manuals set a maximum of 5,000 ppm salt. Beatbot rates the AquaSense 2 for salt water below 5,000 ppm. Check the manual for your model, as product pages rarely state this.

How do I get a heavy robot out of the pool?

Choose a robot with a powered retrieval feature, such as surface parking, wall-climb on command or water release. Nineteen of our 33 records describe one. Lift by the handle, never by a cable.

Can I use an extension lead for the charger or power supply?

The Maytronics Liberty and Zodiac manuals both say not to. Both also call for a 30 mA residual-current device and a 3.5 m gap from the pool edge.

Methodology

Every figure in this guide comes from the RankedRobot database or from a primary source opened while writing. Database records come from maker product pages, spec sheets, stores and user manuals. Each figure carries a source link and a confidence label: maker-stated, reported (from a retailer or third party) or unconfirmed. We flag reported and unconfirmed figures in the text and tables. Each robot page lists its sources.

Our database normalises units to square metres, metres, kilograms and minutes. Prices are shown in the currency and region the maker published, with no currency conversion. Sale and list prices were captured when each record was last checked and change often. Medians and counts in the analysis section use all 33 pool-cleaner records unless a smaller denominator is stated.

For context the database lacks, we read three Maytronics and Zodiac user manuals, maker product pages from Aiper, Beatbot, Maytronics, Roborock, WYBOT and Zodiac, and one peer-reviewed survey on underwater communication. We checked these between 1 and 9 October 2026.

We have not tested any of these robots hands-on. This is spec-based analysis. For how we collect and grade data, see our methodology.

Spotted an error or a newer figure? Email corrections@rankedrobot.com and we'll check it and log the correction.