✅ Worth it? Yes. After 12 months of daily use on its original filters, Cloud RO earned 99/99 in our second Tap Score test. We’d buy it for remineralized RO water, steady stored pressure, battery-powered operation, and app-based filter tracking.
Skip it if cabinet space is tight or you want Wi-Fi monitoring away from home. It’s also a poor fit if you prefer inexpensive standard cartridges sold individually.
- Filtration (RO) 45% 5/5
- Flow/Pressure 20% 5/5
Full scoring basis
- Filtration (RO) (45%): Score: 5/5 – The September 2025 Cloud-spout sample earned 97/99. The August 2026 retest earned 99/99 after 12 months on the original filters. TDS measured 30.8 ppm in the retest, while lead, fluoride, copper, nitrate, and total THMs were below their laboratory reporting limits.
- Flow/Pressure (20%): Score: 5/5 – The 2.8-gallon pressurized tank and battery-powered pump delivered one gallon in 2 minutes 19 seconds in our test. The tank hasn’t run dry during 12 months of use by a family of four.
- Install/Maint (10%): Score: 5/5 – Our installation took under 30 minutes, followed by several hours of flushing. The app tracks filter status by use. The original filters were still installed at the 12-month test, and the current replacement pack includes four cartridges plus a battery.
- Build (10%): Score: 5/5 – The compact filter module uses a separate storage tank, an internal permeate pump, twist-lock cartridges, app monitoring, and no wall outlet. The larger total footprint is the main tradeoff against a tankless system.
- Taste (5%) – Score: 5/5
- Cost (5%) – Score: 5/5
- Certs (5%) – Score: 5/5
- Overall: 5/5 — weighted average of the factors above
We installed Cloud RO in August 2025 after three years with the Waterdrop G3P800. The hardware went in quickly. The initial fill-and-drain sequence took much longer, so this isn’t a project to start an hour before dinner.
Three weeks after installation, water from the Cloud faucet scored 97/99 through Tap Score. We sampled the same faucet again 12 months later, before replacing any filters. That second sample scored 99/99.
What we like
- Original filters still produced a 99/99 Tap Score result at 12 months
- Remineralization keeps the finished water from tasting stripped
- The battery-powered pump doesn’t need an under-sink outlet
- The app shows filter status, tank volume, TDS, battery, and gallons used
- The storage tank kept up with four people in our household
What could be better
- The app uses Bluetooth, so it can’t be checked away from home
- The filter module and tank take more cabinet space than a tankless RO
- Cloud sells the replacement cartridges as a complete $200 pack
- A 1:1 pure-to-waste ratio is less efficient than Waterdrop’s 3:1 ratio
Best for: Households that want tested under-sink RO performance, built-in remineralization, stored pressure, and no electrical outlet.
What changed after 12 months?
Nothing failed, and flow still felt the same. The tank continued to cover normal drinking and cooking demand. The app hadn’t called for a filter change before we collected the anniversary sample.
The better evidence came from the lab. We sampled water directly from the Cloud faucet in September 2025 and again in August 2026. Both samples went through an Advanced City Water Test. The second was collected after 12 months and 204 gallons recorded by the app, with the original cartridges still installed.
✓ QWL long-term lab result – 99/99 Tap Score after 12 months and 204 gallons on the original filters.
| Parameter | September 2025 Cloud spout | August 2026 Cloud spout |
|---|---|---|
| Tap Score | 97/99 | 99/99 |
| Time in service | About 3 weeks | 12 months |
| Filters | Original | Original |
| pH | 7.5 | 7.0 |
| TDS | 66.2 ppm | 30.8 ppm |
| Hardness | 36 ppm | 12.6 ppm |
| Calcium | 2.22 ppm | 0.888 ppm |
| Magnesium | 7.41 ppm | 2.53 ppm |
| Sodium | 4.32 ppm | 6.81 ppm |
| Potassium | 2.46 ppm | 0.331 ppm |
| Total THMs | Below reporting limit | Below reporting limit |
| Lead | Below reporting limit | Below reporting limit |
| Fluoride | Below reporting limit | Below reporting limit |
| Copper | Below reporting limit | Below reporting limit |
| Nitrate as N | Below reporting limit | Below reporting limit |
The number we care about most is the second finished-water result. After a year on the original filters, the Cloud sample still had no Tap Score concerns and received the maximum overall score.
The mineral readings moved between tests. That doesn’t prove the remineralization cartridge gained or lost a fixed percentage of performance. City-water chemistry changes, and the amount added by a post-filter can vary with contact time and cartridge condition.
View the September 2025 Cloud report and the August 2026 Cloud report.
What our test can and can’t prove
The 12-month test tells us what came out of our Cloud faucet. It doesn’t establish a universal removal rate for every Cloud system or every contaminant.
Our home also uses whole-house carbon filtration upstream of Cloud. During the August sampling session, the chlorine test supplied with the Tap Score kit read approximately 1–2 ppm at the untreated basement bypass. The treated kitchen faucet and Cloud faucet both read 0 ppm.
That means we can report 0 ppm chlorine at the Cloud faucet, but we can’t credit Cloud alone for the reduction. The whole-house system had already brought the kitchen reading to zero before the water reached Cloud.
The August laboratory baseline also had THMs below reporting limits. We therefore don’t use that matched sampling event to claim a THM reduction percentage. The 2022 source-water report is useful historical context, but it isn’t a same-day Cloud inlet sample.
Installation and flushing
Our hardware installation took under 30 minutes because the sink already had a usable faucet opening. If you need to drill stone or modify crowded plumbing, plan on more time or hire an installer.
The flush took several hours. Cloud uses a storage tank, and a new membrane produces water slowly at first. We filled and drained the tank three times, then purged the faucet before collecting or drinking the water.
Early batches looked cloudy from trapped air. The cloudiness cleared after the system sat and flushed. That’s normal during startup, but it’s another reason not to rush the process.
The filter module measures 11 by 7 by 16 inches. The separate tank is roughly 9 by 9 by 15 inches. Measure the full arrangement, not just the filter module, especially if the cabinet also holds a disposal or pull-out bin.
Flow and daily use
Cloud filled a one-gallon container in 2 minutes 19 seconds during our timed test. The pressurized tank supplies the faucet immediately, while the membrane refills the tank afterward.
We haven’t emptied the tank during normal use by four people. Filling several large containers back to back can still outrun any storage-tank RO, so high-volume users should compare tank recovery rather than faucet flow alone.
The internal pump also makes refrigerator and ice-maker connections practical. Cloud supplies longer tubing and fittings when needed, but the final result still depends on line length, elevation, and refrigerator requirements.
The app, battery, and replacement filters
The app shows inlet TDS, filtered TDS, remineralized TDS, tank volume, battery status, gallons used, and estimated cartridge dates. Those readings help with trends, but they aren’t a substitute for a contaminant test. A TDS sensor can’t identify lead, PFAS, nitrate, or another individual substance.
The biggest app limitation is Bluetooth. It updates when you’re close to the system, not while you’re traveling. We’d prefer Wi-Fi for remote alerts, though Bluetooth avoids another account-dependent connection in the kitchen.
Cloud’s current replacement pack costs $200 and includes the sediment filter, carbon block, RO membrane, remineralization cartridge, and battery. Cloud says replacement timing averages about one year, but the app bases its recommendation on use and system data.
Our original filters were still in service when we collected the 12-month sample. That’s useful ownership evidence, but it doesn’t mean every household will get the same interval. Feed-water quality and consumption change filter life.
Check Cloud’s current replacement-pack details.
Why we chose Cloud over Waterdrop
Our Waterdrop G3P800 still filtered well after more than three years. We switched because Cloud fit our daily preferences better: no outlet, built-in remineralization, stored pressure, and a quieter setup.
Waterdrop remains the better choice when cabinet space and wastewater efficiency lead the decision. Its tankless body is smaller, and its 3:1 pure-to-waste ratio produces three gallons of filtered water for every gallon sent to the drain. Cloud’s 1:1 ratio produces one filtered gallon per gallon sent to drain.
Cloud is the better fit when you don’t have an outlet, want a pressurized tank, or prefer integrated remineralization. Read our full Cloud versus Waterdrop comparison for the complete decision.
Certification, company, and manufacturing
Cloud RO is listed by NSF under NSF/ANSI 58 for TDS reduction. The public NSF listing also identifies the product as lead-free under the applicable plumbing requirements. Cloud publishes additional third-party test results, but those aren’t the same as model-specific NSF certification claims.
Cloud Water Filter, LLC is a US company based in San Diego. Cloud’s FAQ says it sources components globally and uses a primary manufacturing facility in Suzhou, China. The current NSF listing also identifies the facility in Wujiang, Suzhou, China.
Manufacturing location doesn’t replace product-level evidence. We put more weight on the NSF listing, the published performance data, our installation, and the two Cloud-spout lab reports.
Cloud currently advertises a five-year warranty. Warranty terms, prices, and promotions can change, so verify them before ordering.
Does Cloud RO remove microplastics?
Cloud says the system removes microplastics, and reverse-osmosis membranes are physically capable of rejecting particles far larger than dissolved ions. However, QWL hasn’t run a microplastics panel on Cloud RO.
The current public NSF listing we reviewed shows a certified TDS-reduction claim. We didn’t find a separate microplastics claim in that listing, so we label microplastics as a manufacturer claim rather than a QWL-verified result.
Final verdict
Cloud RO remains our top under-sink RO pick after 12 months. The second lab result matters more than a new-product test because it shows what reached our glass after a year on the original filters.
It isn’t the smallest or most water-efficient system. The app can’t be checked remotely, and the $200 filter pack removes the option to replace only one cartridge. Those are real drawbacks.
We’d still choose Cloud for a kitchen where cabinet space allows it and no outlet is available. If you want a compact tankless system instead, compare the other models in our guide to the best reverse-osmosis systems we tested.
Cloud RO FAQ
Yes, if you want an under-sink RO system with remineralization, stored pressure, app tracking, and no electrical outlet. Our Cloud-spout sample earned 99/99 after 12 months on its original filters. Skip it if cabinet space, Wi-Fi monitoring, or individually sold cartridges matter more.
Cloud claims that it removes microplastics, but QWL hasn’t run a microplastics panel on this model. The public NSF listing we reviewed shows a certified TDS-reduction claim, so we don’t present microplastics removal as a QWL-verified result.
Cloud Water Filter, LLC is a US company based in San Diego. Cloud says its primary manufacturing facility is in Suzhou, China, and the NSF product listing identifies the facility in Wujiang, Suzhou, China.
Cloud says the average replacement interval is about one year, with the app adjusting timing based on use and system data. Our original filters were still installed when the Cloud-spout sample earned 99/99 at 12 months.
No. Cloud uses a replaceable battery pack to power its sensors and internal pump. The current replacement pack includes a new battery.
Yes. Cloud’s internal pump is designed to maintain pressure for refrigerator and ice-maker connections. Actual performance depends on tubing length, elevation, and the refrigerator’s requirements.