The number on the box is clean-air delivery at maximum fan speed, the least interesting fact about owning
a purifier. Most people run that speed for about a week, then decide it is too loud and never go back to
it.
What decides whether a machine is still working in year three is duller: what the next set of filters
costs, whether the room-size claim was quoted on the same basis as everyone else’s, and how it sounds at
the speed you actually leave it on. All three sit inside the five axes averaged here. For a bedroom, the
quietness ordering keeps only two of these five and leads with a
wall-mountable machine that does not place on this page at all.
Room-size claims are not quoted on the same basis
This is the most common way a purifier spec sheet misleads, and a wider pool this year makes it obvious.
One machine in the pool advertises the floor area it clears in one hour — one air change. Another quotes
its coverage at two air changes per hour. Most of the category quotes at four to five, the rate at which a
purifier keeps up with a room instead of catching it once. Recomputed on the common basis, both of those
headline figures shrink by half to three quarters, and both machines land mid-table. The performance axis
uses the recomputed figure, which is why neither appears in the five above.
Reviewer consensus has been making this point about room-size marketing for years. It has not changed the
boxes.
Filter economics are the real price
You buy the machine once and the filters for as long as you keep it. That second purchase is where the
ordering gets decided.
The pool now reaches both extremes. At one end is a welded-steel machine whose carbon bed is rated for
five years; at the other, a sealed cartridge from a single supplier replaced every twelve months. Neither
places in the top five, for opposite reasons. The five-year filter is attached to a machine almost nobody
can live beside, and the twelve-month cartridge is attached to a machine that must be re-filtered on one
company’s schedule, from one company, at whatever that company decides.
What places well is the dull middle. Cartridges in a standard size stocked in general retail, third-party
equivalents that fit, a washable pre-filter that stretches the interval. Filter availability therefore
sits inside the value axis rather than in a footnote, and it is most of the reason the mechanical unit at
position five places at all.
It is also what keeps the machine with the best raw throughput here off the top spot. Its filters are
proprietary, physically large, and replaced as a pair.
Sizing, and the mistake that undoes everything
A purifier rated for a large room, placed in a small one, runs on low and is nearly silent. The same
purifier rated for a small room, placed in a large one, runs flat out permanently, is loud, wears its
filter faster, and still underperforms.
Buying one size up is almost always the right call, and most people do the opposite. The small unit at
position four is a good machine and a common mistake at the same time: it earns its place on inexpensive
filters and a low noise floor, and it earns most of its returns from people who put it in a living room.
Check the coverage figure against your room’s floor area at the speed you intend to run, after correcting
for the basis it was quoted on. That is the whole rule.
HEPA, carbon, and ionizers are three different jobs
A true HEPA stage captures particulates — dust, pollen, dander, smoke particles. Every machine here does
that job, and does it well.
Activated carbon captures gases and odours through a completely separate mechanism, and this is where the
pool spreads out. A few ounces bonded to a pre-filter is the norm, including on the unit leading this
page, whose carbon layer is its one real weakness. One machine in the wider pool carries fifteen pounds of
granular carbon and zeolite, which is a different category of thing and the only reason to tolerate
everything else about it.
Ionizers are the contested third category. They charge particles so they clump and settle, and some
designs produce ozone as a byproduct. Every machine here that includes one lets you turn it off. One of
them comes back from a power cut with it silently switched on again.
Sensors are only worth it if they drive something
An air quality readout is pleasant. An auto mode that raises and lowers the fan from that reading is
useful, because the machine then spends most of its life on a low setting and ramps only when something
happens.
The most capable readout in the pool belongs to a machine that is not in the five above: a purifying tower
with a live display, a genuine auto mode, and the weakest clean-air delivery for the outlay on this page. A
sensor does not rescue a throughput deficit, and an average across five axes says so plainly.
The reverse holds too. The entry at the top has no sensor, no display, and one button.
Smart features stay outside the scoring
Smart home integration is not an axis. It changes with firmware, it varies by which ecosystem you run, and
a purifier that needs an app to change fan speed is worse than one with a button.
Filter life indicators are also unscored. They are timers, not sensors, and they count elapsed hours
instead of actual loading.