Buying guides

Home-fit finder

Reviews & explainers

Home-fit finder

Buying guides

Reviews & explainers

Dreame D20 Plus robot vacuum in its white self-emptying dock

ReviewDreame

Dreame D20 Plus review: top carpet pickup, no eyes for obstacles

Verdict

Swept's call: buy the Dreame D20 Plus for carpet, rugs and shedding pets in a home where someone picks up cords and socks before it runs. It can't see obstacles, and its mop is barely worth fitting.

under $300Check price on Amazon (opens in a new tab)

Opens Amazon.com · Prices change often

On the carpet test that separates robots best, the Dreame D20 Plus pulls up more buried grit than almost anything Vacuum Wars has run, whatever it cost. Then it drives straight over your phone charger. Those two facts are the whole robot. That lab ground sand deep into medium-pile carpet, and this robot got 93% of it back out. It collected 91% of the pet hair pressed into the pile, and wrapped none in the lab's long-hair tangle test. It also avoided 0 of 24 objects on the same lab's obstacle course. There's no camera and no carpet sensor, and the mop pad just drags. Swept's call: buy it if your floors are mostly carpeted, a pet sheds on them, and nobody wants a mop. Someone has to clear socks, cords and toys before every run.

Dreame D20 Plusunder $300 Check price on Amazon (opens in a new tab)

BuyCarpet and pet hair on a budget, floor picked up first

Dreame D20 Plus: what the box promises and what you get

Price band: under $300

This is a lidar robot with a bag dock that empties its bin and does nothing else. Dreame sells it on 13,000 Pa of suction. I read that the way I'd read a supplier quoting a part's strength with no test certificate behind it. Two labs measured its suction at the floor and found it below average. Both then watched it clean carpet superbly. The twin rollers do the work. Dreame deserves real credit for the brushes. The pascals are decoration.

Will it keep up with a shedding dog or cat?

Yes, and this is where it earns its place. Its pet-hair pickup beat Vacuum Wars' average. The average robot in that lab's long-hair test wrapped about a quarter of the hair; this one wrapped none. Dreame's detangling promise holds on the roller. Hair still finds other places to go. One tester saw it wind around the wheels, and a long-term write-up warns it can build up in the channel to the dock. Only one lab has weighed its pet-hair pickup, so the result is strong but not yet settled.

Dreame claims up to 150 days per dock bag, by its own calculation. Wirecutter estimates 2–3 months. With a shedder, believe Wirecutter. Big homes are fine: Vacuum Wars projects about 2,761 sq ft on a charge, from a run over part of the battery.

For a shedding dog or cat on carpet, on a tight budget, this is the job we'd buy it to do.

Where it falls down: it cannot see

Will it run over dog poop, cords and socks?

Yes. It finds a sock the way a bowling ball finds a pin. It has a bumper, a low-obstacle sensor and lidar for mapping, but nothing that recognizes what's on the floor. Wirecutter found it prone to getting trapped, and needing more supervision than pricier robots. Nobody tested it with pet mess. But a robot that steered around none of 24 everyday objects gives no reason to hope. If your dog has accidents indoors, this is the wrong robot.

If you've read that its avoidance is middling, check which D20 that review covers. One write-up that says so names the D20 Pro Plus, which has a better front sensor.

Will it get over the step into the kitchen?

Only a low one. The manual caps it at 0.8 in (20 mm). Flat doorways and ordinary rug edges are fine; a taller sill or patio door track stops it. If a step in your house is taller than that, the Roborock Saros 20 is the climber our vacuum-only roundup points to. In one home, the owner reports the D20 Plus fighting the same loose rug run after run. It stands 3.8 in tall with a fixed turret, so measure under the couch.

The mop is an afterthought

The box says "Robot Vacuum and Mop". In practice it's a vacuum that owns a damp cloth. The pad doesn't lift, and the dock doesn't wash it. You fill a 12 fl oz (350 ml) tank by hand. On Vacuum Wars' dried coffee and grape juice stains it scored 90, against that review's average of 107.8. Every tester we read treats the mop as secondary. With no carpet sensor, the robot doesn't know a rug is a rug either. Testers drew carpets in the app or took the pad off.

If you never wanted a mop, our vacuum-only roundup is the page for you. If you do want the wood between your rugs mopped, and those rugs must stay dry, the Roborock Qrevo Curv 2 Flow is the robot our carpet roundup recommends for rugs laid over hard floors. Where the rugs are deep, the pet sheds hard and baseboards matter most, our mop roundup turns to the Dreame X60 Max Ultra Complete instead.

Living with it: noise, app and warranty

Noise is a split. Vacuum Chef measured 74 dB at full suction, slightly above average. A tester who bought his own called it very quiet. Nobody has measured the dock emptying.

The buttons start a clean, a spot clean and send it home. Maps, rooms and schedules need the Dreamehome app and an account. The manual says it connects to 2.4 GHz Wi-Fi only. There's no Matter support and no Home Assistant route.

Dreame covers the robot and dock for 3 years, and the battery for one. Its after-sales policy limits that cover to Dreame's authorized US channels, so check who is selling it. Nobody has reliability data beyond about three months, and no recalls were found.

Does it have a camera or microphone?

Neither, per the US manual's parts list. The manual also gives a Wi-Fi off switch: hold Dock and press Power 10 times. Wi-Fi turns off once the robot is fully charged, and any button turns it back on. A Dreamehome app flaw once exposed logins on untrusted networks, this robot's included. CISA says later app versions fix it. We found no US security-update period from Dreame, so how long fixes keep coming is unknown. It went on sale before the FCC listed foreign-made robots in July 2026, so you can still buy it and run it legally.

If the missing camera is what draws you, our guide to robots without a camera or microphone weighs it against the alternatives. Where the floors are hard with low rugs, the shedder is house-trained and you want it mopped too, that guide turns to the eufy Omni C28 instead. A different buyer wants years of patches and control without the cloud. This one offers neither a dated US promise nor any local control. For that buyer, our guide to the FCC question settles on the Roborock Saros 20, the same climber named above.

Keep the four D20s apart

Which Dreame D20 am I looking at?

Four Dreame budget robots share the name. Only the first row is the robot this review measures.

RobotModel codeHow it differsMeasured on this page
Dreame D20 PlusRobot RLD35GD, dock RCDD0301This reviewYes
Dreame D20 Pro PlusRLD43SDAdds an extending side brush, an ultrasonic carpet sensor and a better front obstacle sensor; same bin and tankNo
Dreame D20 Air PlusRLF41GDCheaper, newer, with a higher suction claimNo
Dreame D10 Plus Gen 2Not givenAn older generationNo

Dreame's own D20 Plus page muddles things. Its prose promises an extending side brush. Its spec table says there isn't one, and testers agree. The sentence appears to have wandered over from the Pro Plus. On paper the Pro Plus fixes two of this robot's weaknesses. No lab in our evidence has measured whether it does, so we can't call it the better buy.

What we don't know yet

Nobody has tested it on high pile, shag or rug fringe. Only one lab has weighed its pet-hair pickup. Nobody knows how often the dock path clogs with a heavy shedder, or how it holds up past a few months. Repeated owner reports of clogs, or poor results on thick carpet, would change our call.

Dreame D20 Plus robot vacuum in its white self-emptying dock

BuyCarpet and pet hair on a budget, floor picked up first

Should you buy the Dreame D20 Plus?

Buy it for a carpeted home with a shedding dog or cat, on a tight budget, where nobody needs a mop and someone tidies first. There it vacuums like a much dearer robot, and it doesn't tangle. The budget is what separates its buyer from the next one up. If you can spend more and want a robot that sees what's on the floor, the Dreame L60 Ultra PE is our choice for carpet and a shedder that doesn't have accidents, in a house whose doorways are under an inch; it has cameras, and our carpet roundup names it for wall-to-wall carpet on one floor. If your pet sheds hard or has accidents indoors, look at the Dreame X60 Max Ultra Complete, which leads that same roundup for carpet-heavy homes. If your floors are mostly hard, your pet sheds an ordinary amount and you want them mopped, our mid-budget roundup prefers the ECOVACS DEEBOT T90 Pro Omni. Still unsure? Start with the home-fit finder, or with whether a robot is worth having at all.

Pick up the floor, and it will do the rest.

under $300Check price on Amazon (opens in a new tab)

Opens Amazon.com · Prices change often

The evidence in full

17 sources

Vacuum Wars (review dated 2026-10-21)

Vacuum Wars buys its robots and tests at maximum power. Its review robot matches this variant.

TestD20 PlusVacuum Wars' average on the same test
Carpet deep clean (sand embedded in medium-pile carpet)93%77.86
Flattened pet hair on carpet, two runs91%82.2
Long-hair tangle (share of hair wrapped)0%26
Obstacles avoided0 of 2415.7
Dried-stain score (coffee and grape juice)90107.8
Crevice test11.97
Suction at the brush housing0.53 kPa0.99
Airflow18 CFM16.95
Navigation speed0.75 m2/min0.711
Battery pace, in minutes of cleaning per percent of battery (measured over part of the battery, not a rundown)3.421.56
Estimated coverage per charge (projected)2,761 sq ft1,170

Vacuum Wars confirms the pad doesn't lift and the dock doesn't wash it or refill water. Its verdict: a vacuum-first robot for uncluttered homes.

Vacuum Chef (review dated 2026-05-25)

Review robot sent by Dreame; the review says it is unsponsored, and its images match this variant. It measured 74 dB at maximum suction, slightly louder than average; in-use carpet airflow of 383 FPM against its average of 423; and in-use pressure of 0.18 kPa, below average. Its real-world cleaning verdict was strongly favorable despite those below-average power readings. It noted small or low obstacles may be bumped or tangled, and that the tank takes water only, no detergent. It publishes no pet-hair figure.

Wirecutter (combo guide, 2026)

Estimates the bag lasts 2–3 months against Dreame's 150-day claim. Calls it prone to getting trapped by cords, rugs and objects, which is entrapment, not hair on the roller. Rates its mopping as light wiping only. Reports a three-year parts warranty with one year on the battery, and a dock noticeably narrower than its other picks.

Just A Dad Approved

The reviewer bought his own robot. His first-look video, dated 2025-09-07, is the earliest dated video of this model we found, and in it he calls the robot brand new. He found no camera and no carpet sensor, with a low-obstacle front sensor and a side brush that doesn't extend. By ear he judged the robot very quiet at maximum and decent with long hair, leaving a few hair balls; its mop "just drags". His side-by-side video confirms only the Pro Plus has the extending brush, carpet sensor and better front sensor. His carpet video (kitty litter, long hair, ground coffee) found the auto-empty did a good job, though the filter was starting to clog and long hair wrapped around the wheels. His written three-month review scores it as a vacuum-only robot and advises leaving the pad off, rates avoidance "mid-tier" and warns the auto-empty path can clog with hair; its text names the D20 Pro Plus, so we treat it as possibly describing the other variant.

Where sources disagree

  • Obstacles: Vacuum Wars' 0 of 24 against a "mid-tier" pass in the written review. Explained by method (a fixed course of low objects against a real room with larger items) and by the written review naming the Pro Plus.
  • Bag life: Dreame's 150 days is an in-house calculation; Wirecutter's 2–3 months reflects real debris.
  • Suction: the 13,000 Pa claim against below-average measured suction at both labs. Both labs note measured suction did not predict its strong carpet results.
  • Side brush: Dreame's page prose says it extends; the spec table and testers say no.
  • Noise: "very quiet" by ear against 74 dB measured, slightly above average.

Identity, safety and security

  • Robot RLD35GD with dock RCDD0301, per the US manual (edition US-S00, made in Vietnam). On sale in the US from September 2025. The manual prints a Part 15 statement and no FCC ID; the FCC says devices authorized by declaration carry none and still count as authorized, and that previously authorized robots can still be sold, get basic firmware updates and be repaired with identical parts.
  • CPSC recall search for Dreame returned nothing on 2026-10-26.
  • No camera or microphone in the US manual's parts list. The 2024 SEC Consult finding (shared users could list and delete the owner's robot photos and videos) concerns camera models and doesn't apply to this robot's hardware. The 2025 app TLS flaw (CVE-2025-8393) did reach it through the app; CISA's update says later app versions enforce certificate checks.
  • We found no published security-update period from Dreame covering US robots.
  • No Matter claim, no official Home Assistant integration, and not on Valetudo's supported list.
  • Owners: one Reddit thread on this exact model, about a loose rug; its owner's location and time owned aren't stated.

Gaps

Noise is only half measured: nobody has recorded the dock emptying or the robot in quiet mode. Carpet is only tested at medium pile, with no high-pile, shag, fringe or dark-carpet results. Long-term ownership is a blank beyond about three months, parts availability included, and so is the dock's clog rate in heavy-shedder homes. Dreame doesn't name its obstacle sensor, and only one lab has weighed pet-hair pickup. RTINGS, Tom's Guide, TechGearLab and CNET hadn't reviewed it when we checked.

What we read

Research count for this page: 1 models weighed; 1 maker manuals among them; across this site's research, 497 sources kept in full; 148 Reddit threads read, kept as the reader's summary, not in full.