A strong mid-price pick for pet homes with carpet: it measured near the top on sand and hair and climbs thresholds most robots at this price can’t.
Amazon $759.99The Dreame L50 Ultra is the better robot vacuum overall, rated 4.4 to 3.9. The Dreame L50 Ultra leads on cleaning, mopping and navigation. The Dreame L50 Ultra costs $10 less ($759.99 vs $769.99).
Fine for smaller homes with mostly hard floors and carpet that needs a deep clean, but the newer ThermaCharged adds heated washing and better pet hair pickup for less money at Shark.
Amazon $769.99Prices were last checked October 26, 2026 and are subject to change; the price and availability shown by the retailer at the time of purchase apply.
Which should you buy?
- Choose the Dreame L50 Ultra for cleaning (measured 90% of sand from carpet, 100% of pet hair, no hair tangled vs measured 90% of sand from carpet, 83.2% of pet hair, some hair tangled (0.3)), mopping (Dual spinning pads, leaves its mop at the dock for carpet, extends to edges vs Vibrating pad, leaves its mop at the dock for carpet, extends to edges) and navigation (LiDAR + camera vs LiDAR + 3D light).
Scores side by side
Each factor is scored 0-100 against the same fixed reference points for every self-cleaning dock robot vacuum; a clear win is highlighted. How we rank explains each one.
| Factor weight | Dreame L50 Ultra | Shark PowerDetect NeverTouch Pro |
|---|---|---|
| Overall rating | 4.4/5 | 3.9/5 |
| Cleaning 30% | 88 measured 90% of sand from carpet, 100% of pet hair, no hair tangled | 73 measured 90% of sand from carpet, 83.2% of pet hair, some hair tangled (0.3) |
| Value 25% | 53 $760 | 52 $770 |
| Dock 20% | 100 Self-empties the bin, washes the mop, hot-water mop wash, hot-air mop drying, doses detergent | 76 Self-empties the bin, washes the mop |
| Navigation 15% | 100 LiDAR + camera | 85 LiDAR + 3D light |
| Mopping 10% | 86 Dual spinning pads, leaves its mop at the dock for carpet, extends to edges | 62 Vibrating pad, leaves its mop at the dock for carpet, extends to edges |
Independent tests
Dreame L50 Ultra
| Tested by | Carpet deep clean | Pet hair | Hair tangle (0 = none) | Area per charge | Conditions |
|---|---|---|---|---|---|
| Vacuum Wars | 90% | 100% | 0 | 1,129 sq ft | Embedded sand in medium-pile carpet; flattened 2.5 in pet hair on carpet, two runs; 7 in hair tangle on hard floor (0 = none wrapped). |
Shark PowerDetect NeverTouch Pro
| Tested by | Carpet deep clean | Pet hair | Hair tangle (0 = none) | Area per charge | Conditions |
|---|---|---|---|---|---|
| Vacuum Wars | 90% | 83.2% | 0.3 | 698 sq ft | Embedded sand in medium-pile carpet; flattened 2.5 in pet hair on carpet, two runs; 7 in hair tangle on hard floor (0 = none wrapped). |
Specifications
| Dreame L50 Ultra | Shark PowerDetect NeverTouch Pro | |
|---|---|---|
| Vacuum | ||
| Claimed suction | 19,500 Pa | - |
| Main brush | HyperStream Detangling DuoBrush (bristled rubber brush plus TPU rubber brush) | - |
| Dustbin | 395 mL | 284 mL |
| Extending side brush | Yes | - |
| Mopping | ||
| Mop | Dual spinning pads, leaves its mop at the dock for carpet, extends to edges | Vibrating pad, leaves its mop at the dock for carpet, extends to edges |
| Mop lift | 10.5 mm | - |
| Leaves mop at dock | Yes | Yes |
| Extending mop | Yes | Yes |
| Dock | ||
| Dock | Washes mop + empties | Washes mop + empties |
| Self-empties | Yes | Yes |
| Dust bag | 3.2 L | - |
| Washes mop | Yes | Yes |
| Hot-water wash | Yes | No |
| Hot-air drying | Yes | No |
| Detergent dosing | Yes | - |
| Clean water tank | 4.5 L | - |
| Navigation & app | ||
| Navigation | LiDAR + camera | LiDAR + 3D light |
| Obstacle avoidance | 3D structured light plus AI RGB camera, recognises 180+ object types | 3D sensor obstacle detection |
| Climbs thresholds | 60 mm | - |
| Matter | Yes | No |
| Voice assistants | Alexa, Siri, Google Home | Alexa, Google Assistant |
| Build | ||
| Height | 3.8 in (97 mm) | - |
| Battery | 6,400 mAh | - |
| Run time | Up to 3 h 20 min | Up to 2 h |
| Weight | 9.8 lb (4.4 kg) | - |
| Warranty | 3 years | 3 years |
| Released | 2025-02 | 2024-09 |
