We have not used any of the products on this site. Every recommendation is arrived at by taking the figures the manufacturer publishes, applying arithmetic that is printed in full on this page, and saying plainly where a figure does not exist. If you disagree with a pick, you can find the exact step where we diverge.
That is an unusual thing for a gear site to say out loud, so it is worth being direct about why. The ranking sites in this category lead with test labs and credentialed reviewers. We have neither. What we could do honestly is the thing almost nobody in this niche does: publish the calculation. A beginner standing in a shop does not need to know how a filter felt to us. They need to know whether the number on this box clears the number their tank needs, and that is a question arithmetic answers better than an opinion does.
Rule one: every number has a document behind it
Every figure in a specification table on this site is quoted from the manufacturer's own product page, a manual, or a retailer listing that states the contents. Those links sit at the foot of every page that uses them, not in an appendix. Where a maker does not publish a figure — and this is common, particularly for bundled kit filters and air-driven sponge filters — the page says the figure is not published rather than estimating one.
That rule costs us pages. There are comparisons we would like to have written and could not, because one of the two products has no published flow rate and a comparison built on a guess is worth less than no comparison at all.
Filter sizing: turnover, and the 30% you do not get
Filters are sold on a tank-size label — "for tanks up to 20 gallons" — and that label is the least useful number on the box, because it does not tell you what stocking or what media load it assumes. The figure that means something is turnover: how many times per hour the filter moves the tank's whole volume.
- A lightly stocked or planted tank wants roughly 4 to 6 times turnover per hour.
- A typical beginner community tank — which is what most readers here are building — wants roughly 6 to 8 times.
- We size to 6x as the working default, because it suits the stocking level a first tank realistically ends up at.
So a 10 gallon community tank wants about 60 gallons per hour of real flow, and a 20 gallon wants about 120.
Then the correction almost nobody applies
A filter's rated GPH is measured under conditions you will never reproduce: brand-new empty media, and no height to lift the water back into the tank. In service, loaded with sponge and biomedia and pumping water up over the rim, real throughput is commonly 25 to 40% below the rating, and it falls further as the media loads with detritus between cleanings.
We assume a 30% loss. That means the rated figure you should shop for is the flow the tank needs divided by 0.7:
This single step is why our filter picks differ from most lists. A filter rated at exactly the flow your tank needs is under-filtering in practice, and a filter labeled "for 20 gallon tanks" quite often does not clear 170 rated GPH. We check that explicitly on every filter page and print the result.
| Tank | Real flow needed at 6x | Minimum RATED GPH to shop for |
|---|---|---|
| 5 gallons | 30 GPH | 45 GPH |
| 10 gallons | 60 GPH | 85 GPH |
| 20 gallons | 120 GPH | 170 GPH |
| 29 gallons | 174 GPH | 250 GPH |
| 40 gallons | 240 GPH | 345 GPH |
Computed with the formula above. Round up, never down: over-filtering a freshwater community tank is close to harmless, and the fix for too much current is a baffle or a spray bar, which costs nothing.
Heater sizing: wattage is about the room, not just the tank
The rule of thumb printed on most packaging — 5 watts per gallon — is a single number standing in for two variables, and it is why so many beginners end up with a heater that cannot hold temperature in January. Wattage is a function of the tank volume AND the lift: how far above your coldest realistic room temperature you need to hold the water.
| Lift needed | Watts per gallon | What that looks like |
|---|---|---|
| 5 °F | 3 W/gal | A warm room that barely drops overnight |
| 10 °F | 5 W/gal | The common case, and the source of the 5 W/gal rule |
| 15 °F | 7.5 W/gal | A cool room, or a house that drops hard at night |
| 20 °F | 10 W/gal | A cold basement or an unheated room in winter |
Lift is measured against your COLDEST realistic room temperature, not the average. A heater sized to the average fails on the coldest night, which is exactly when failure matters.
Worked: a 20 gallon tank held at 78 °F in a room that reaches 68 °F overnight needs a 10 °F lift, so 20 x 5 = 100 W. The same tank in a 58 °F basement needs a 20 °F lift, so 20 x 10 = 200 W — double, for an identical tank. Any list that recommends a heater by tank size alone cannot tell you which of those you are.
One consequence worth stating: a preset heater, which holds whatever temperature its maker chose and gives you no dial, is fine in the first case and wrong in the second. Several popular starter kits ship exactly that, which is why we flag it on every kit page.
Why we will not publish a score
No product on this site carries a rating out of ten, a star count, or an aggregate score, and none of our structured data claims one. A score is a summary of testing. We have not tested these products, so a score from us would be a summary of nothing, dressed up to look like measurement.
What we publish instead is the figure each pick was judged against — the required GPH it clears or misses, the lift its wattage can hold — plus a plainly worded verdict and, on every roundup, at least one product we tell you not to buy first and why. A shortlist where everything is recommended is a catalog.
What we deliberately do not do
- No invented specifications. If a maker does not publish a flow rate, we do not print one. Air-driven sponge filters are the recurring example: output depends on the air pump you pair with them, and no manufacturer publishes a meaningful GPH.
- No testimonials or user reviews. We do not have verified first-hand accounts to publish, and inventing them is the fastest way to make everything else on the site worthless.
- No urgency, no fake scarcity. No countdowns, no "prices rising", no "only a few left". An aquarium purchase should be slower than the shop wants it to be, not faster.
- No commission-led rankings. Amazon pays us a percentage. A brand-direct store might pay more. Where a brand-direct link would be better for you we say so, and where it would be worse we send you to Amazon anyway.
- No stale prices. Prices render only from the live price layer and expire after 48 hours. When there is no live figure the button reads "Check price" instead of showing a number that may be months old.
Where the arithmetic came from
The turnover multiples and the real-world throughput correction are long-standing hobby practice rather than a published standard, and we would rather say that than dress them up. What is documented is everything they are applied to: the rated flow figures, the heater wattages and tank ranges, and the test-kit parameters all come from the manufacturer pages linked below. Ammonia toxicity rising with pH and temperature — the mechanism behind the whole cycling argument — is documented by the EPA, also linked below.
If you think a step here is wrong, the useful thing about publishing it is that you can say which step. That is the point. Now go and use it: how to size an aquarium filter, or start at the beginning with start here.