Air Purifiers · Guide
HEPA vs Ionizer vs UV-C: Which Air Purifier Technology Actually Cleans Your Air
Air purifier marketing is a competition to name technologies. True HEPA, H13, H14, activated carbon, ionizers, PlasmaWave, UV-C, plasma, photocatalytic oxidation - each arrives with a claim and a badge on the box. Very few of them change how much clean air reaches your lungs, and a couple of them can make the air worse. This guide takes the stages one at a time and asks what each one physically does, what it costs, and whether the catalog's own numbers support the claim. The through-line is simple: particle filtration is the mature, proven technology, and everything else is either a genuine supplement for a specific problem, a marginal extra, or a feature you should switch off. The scores quoted throughout come from a deterministic engine reading published specifications - AHAM CADR figures, coverage, noise floors, wattage - not from lab testing. Where a well-known machine scores badly, the reason sits in its numbers, and it is stated plainly.
How mechanical filtration actually works, and the 0.3 micron myth
A HEPA filter is not a sieve. It is a dense mat of tangled fibers, and it captures particles in several ways at once: large particles slam straight into fibers, medium ones brush past and stick, and the smallest ones bounce around erratically in the air stream and collide with a fiber almost by accident. That last mechanism, Brownian diffusion, is why the most common worry about HEPA is backwards. Buyers read "removes 99.97% of particles at 0.3 microns" and conclude that anything smaller slips through. The opposite is true. 0.3 micron is the most penetrating particle size - the worst case, the size where impaction has faded and diffusion has not yet taken over. Standards test at that size precisely because it is the hardest one to catch. Particles smaller than it, including most combustion and wildfire smoke particles, are captured at higher rates, not lower. What this means at the point of purchase: filter efficiency is almost never your limiting factor. Nearly every unit in this catalog runs True HEPA media, from the $90 GermGuardian AC4825E to the $530 Dyson Purifier Cool TP07. What separates them is how much air they push through that media per hour, whether any of that air can sneak around it, and whether you will actually run the machine at the speed its rating assumes.
True HEPA, H13 and H14: what the grades mean and why they rarely decide anything
True HEPA is a US convention: media that removes 99.97% of particles at 0.3 micron. H13 and H14 come from the European EN 1822 classification, which tests each filter at its own most penetrating particle size and sets the bar at 99.95% for H13 and 99.995% for H14. On paper, H14 looks like a meaningful step up. In a room, it is close to a rounding error. Run the arithmetic on what escapes. True HEPA lets roughly 3 particles in 10,000 through per pass. H14 lets through about 5 in 100,000. Both are so far past the point of diminishing returns that the number of passes matters vastly more than the grade of the media. A purifier that cycles a room's air five times an hour at 99.97% leaves the air far cleaner than one cycling it twice at 99.995%. The catalog makes the trade-off concrete. The Medify MA-40 is the H14-grade option here: $290, 224 CADR, 420 sq ft of AHAM coverage. The Coway AP-1512HH Mighty is ordinary True HEPA and delivers 233 CADR over 361 sq ft for $170. More air, more cheaply, at the lower grade. The MA-40 scores 5.0 overall against the Coway's 7.0, and the filter grade does nothing to close that gap.
Sealed systems and the air that never touches the filter
Grade only describes the media. What matters just as much is whether all the air is forced through it. Inexpensive housings leak: air takes the path of least resistance around a loosely seated filter, through gasket gaps, or past an ill-fitting door, and that bypassed fraction is never filtered at all. This is what "sealed system" marketing is genuinely selling, and it is a real engineering cost. It is also what the Dyson Purifier Cool TP07 asks you to pay for. At $530 it is the most expensive unit in the catalog, built around a sealed H13 system with sensors, an app, and an oscillating bladeless fan. Dyson publishes no CADR at all, and independent measurements place its raw throughput below same-price towers, which is why it scores 5 out of 10 on clean air - below the $170 Coway and the $220 Levoit Core 400S, which both score 6 - even while scoring a perfect 10 on smart features. Rated coverage is 300 sq ft on 40 W. The entry Purifier Cool Gen1 TP10 is the same format without Wi-Fi, at $380 for 250 sq ft, scoring 5.7 overall. Neither is a bad machine. But a premium sealed filter inside a modest fan still yields a modest volume of clean air, and the TP07's 7.7 overall is carried by sensors, app and fan duty rather than by filtration throughput.
Activated carbon: the only stage that touches odors and VOCs
HEPA media does nothing about smells. Odors, cooking fumes, the gas phase of wildfire smoke and volatile organic compounds are molecules, not particles, and they pass through a particle filter untouched. Removing them requires adsorption: activated carbon with an enormous internal pore surface, where gas molecules stick to the pore walls. The variable nobody prints on the box is how much carbon is actually in there. A thin sheet of carbon-impregnated fabric wrapped around a HEPA cylinder holds a few grams and saturates within weeks. A dedicated pelletized bed holds far more and lasts far longer. When a listing calls carbon a "layer" rather than a filter, assume the former. This is where the Winix 5500-2 earns its keep at $160: a washable AOC carbon filter that is a separate stage rather than a coating. Be honest about what washing does, though. Rinsing clears surface dust and restores airflow. It does not regenerate carbon that has already adsorbed its fill, because regeneration takes industrial heat. Treat washing as maintenance, not renewal, and replace the carbon when odors stop improving. Two limits are worth knowing. Plain carbon handles formaldehyde poorly, which is why impregnated media exists. And nothing here touches carbon dioxide, so a stuffy, poorly ventilated bedroom will still feel stuffy after purification. That is a ventilation problem, and no purifier solves it.
Ionizers and PlasmaWave: two different machines under one word
"Ionizer" covers two very different designs, and the difference decides whether it is useful or a liability. The first charges particles inside the airflow, immediately upstream of a filter that then collects them. Charged particles clump and adhere to media far more readily, so the filter can be less dense, which means less resistance, less noise and less power for the same throughput. Blueair's HEPASilent works this way, and the numbers show what it buys: the Blue Pure 211i Max moves a 400-class CADR across 635 sq ft on 45 W at a 23 dB low setting, and scores 9 on clean air. Nothing else in this catalog combines that throughput, wattage and noise floor. The second design emits ions into the room. Particles pick up a charge and settle onto walls, curtains and floors. They leave the air, which a sensor may register as improvement, but they are not removed from the home - they are redistributed, and they re-enter the air when you move around or vacuum. Ion emitters can also drive reactions with common indoor chemicals, particularly the terpenes in cleaning products and air fresheners, forming formaldehyde and fresh ultrafine particles. Winix's PlasmaWave belongs to this second family and appears on both the 5500-2 and the 9800. The detail that matters on the 5500-2 is that it is switchable. Use the switch.
UV-C claims versus contact-time reality
UV-C germicidal claims are the least reliable marketing in this category, and the reason is dose. Inactivating a microorganism requires a specific ultraviolet energy dose, and dose is intensity multiplied by exposure time. A consumer purifier gives you very little of the second term. Do the arithmetic roughly. Air enters the machine, passes a small lamp inside a chamber a few inches deep, and leaves. At the flow rates these units run - the GermGuardian AC4825E is rated around 99 CADR, the Medify MA-40 around 224 - each parcel of air spends a fraction of a second beside that lamp. Purpose-built germicidal systems compensate with high-output lamps, reflective chambers and deliberately slowed airflow. A tower on your floor does none of that, and a lamp powerful enough to compensate brings problems of its own. Meanwhile the HEPA media sitting beside the lamp is already capturing bacteria, mold spores and virus-laden droplet nuclei mechanically, in a single pass, with no dose requirement whatsoever. Physically trapping a pathogen and inactivating it are equivalent outcomes if the air no longer carries it. The GermGuardian AC4825E is the clearest warning in the catalog. At $90 it leads with UV-C, and it scores 3.0 overall - the lowest here - on 99 CADR, 153 sq ft of coverage and a 40 dB noise floor. The UV badge bought no clean air.
Ozone, certification, and the stages you should switch off
Ozone is the reason to read the feature list rather than the headline. Ion emitters and certain UV lamps can generate it as a by-product, and ozone is a lung irritant rather than an air cleaner - it is regulated as a pollutant outdoors for exactly the reason it is unwelcome in a bedroom. It also reacts with indoor chemicals to produce formaldehyde and ultrafine particles, so a machine that makes ozone can measurably worsen the air it was bought to fix. Two checks are worth the minute they take. California's Air Resources Board maintains a list of air cleaners certified for sale in the state, which is the strictest widely available screen for ozone emission. Separately, UL 2998 is a validation specifically for negligible ozone output. If a unit with an ionizer appears on neither, treat that as a reason to look elsewhere. Then the practical rule: switch off anything that is not filtration. Ionizers, PlasmaWave and UV lamps are optional stages on every unit here that offers them, and disabling them costs almost nothing in particle removal while removing the by-product risk entirely. One caveat is worth knowing. AHAM measures CADR on the machine as sold, so if a unit ships with its ionizer active, a small share of the rated figure may come from it. In a bedroom, that trade is still worth making.
CADR is the number that survives all of this
Strip away the technology names and one number decides how clean a room actually gets: CADR, the Clean Air Delivery Rate, expressed in cubic feet per minute of particle-free air delivered. It already folds filter efficiency and airflow into a single figure, which is precisely why manufacturers with weak numbers would rather talk about filter grades and lamps. Ranked on clean air delivered rather than on features named, the catalog reorders itself. The Blueair Blue Pure 211i Max delivers the most outright: a 400-class CADR across 635 sq ft for $350, scoring 8.7 overall. On CADR per dollar the workhorses win - the Winix 5500-2 at 232 CADR for $160 and the Coway AP-1512HH Mighty at 233 CADR for $170 both beat it, and the Levoit Core 400S returns 256 CADR at $220. Against those, the $290 Medify MA-40's 224 CADR and the $530 Dyson TP07 look expensive per unit of clean air, and their 5.0 and 7.7 overall scores reflect where the money went. Watch the coverage claims as well. Manufacturers quote two different numbers: an AHAM-derived figure and a much larger square-feet headline based on a single air change per hour. The Honeywell HPA300's 2,250 sq ft claim is the one-air-change figure; its AHAM coverage is 465 sq ft. The Medify MA-40's pair is 1,793 sq ft and 420 sq ft. Always compare the smaller number.
Noise and wattage are filtration specs in disguise
A CADR rating is measured at maximum speed. If a machine is too loud to run at maximum, you never receive the number you paid for, which makes the noise floor a filtration spec rather than a comfort extra. The catalog's floors run from 18 dB to 40 dB, and that range decides which units survive contact with a real bedroom. The Honeywell HPA300 is the clean illustration. It has genuine throughput, 300 CADR over 465 sq ft, but delivers it at a 32 dB floor and 130 W, with manual controls and no auto mode. It scores 5.3 overall despite a 7 on clean air, and a 4 on smart. It is a machine you run hard while you are out of the room. The GermGuardian AC4825E's 40 dB floor is worse and comes with only 99 CADR to show for it, scoring 2 out of 10 on quiet. At the other end, the Blueair Blue Pure 411i Max runs at 18 dB on 10 W, the quietest and lowest-draw unit here, scoring a perfect 10 on quiet - but its 141 CADR and 219 sq ft make it a bedroom machine and nothing else, and it scores 4 on clean air. The Coway AP-1512HH Mighty's 24 dB floor alongside 233 CADR is the most usable middle ground on offer.
Match the technology to the problem you actually have
Different problems reward different stages, and buying the wrong stage is the most common mistake in this category. Pollen and dust are pure particle problems. Carbon, ionizers and UV add nothing to them. All you need is enough CADR for the room and a machine quiet enough to leave running overnight through the season - the Levoit Core 400S at $220 with 256 CADR over 433 sq ft, or the Coway at $170, cover this completely. Pet dander is also particulate, but hair and heavy fur clog HEPA media quickly and make replacement filters the real running cost. A washable pre-filter is the feature that matters, which is why the Levoit Vital 200S at $190 is built the way it is: 250 CADR smoke, 380 sq ft, and a pre-filter you rinse instead of replace. Smoke and cooking odors are the one case that genuinely needs two technologies at once - throughput for the particles, carbon for the gases - which points at a unit with a real separate carbon stage. Household chemicals, new-furniture off-gassing and formaldehyde are the hardest case, and the honest answer is that consumer carbon helps modestly while ventilation helps more. Finally, size the machine to one room. Purifiers clean the room they stand in, with the door closed. Two mid-priced units in the two rooms you occupy beat one expensive unit in the hallway.
What this adds up to when you buy
Put the technologies in order of what they reliably deliver and three buying positions fall out. If you want filtration and nothing else, the Coway AP-1512HH Mighty is the cleanest expression of it: 233 CADR smoke, 361 sq ft, a 24 dB floor, and a sensor-driven auto mode with an eco setting, at $170 and no app. It scores 7.0 overall with an 8 on quiet, and there is no ionizer or lamp on it to think about. For most buyers, most of the time, that is the right amount of technology. If odors are the problem, the Winix 5500-2 at $160 is the case for buying carbon deliberately: a separate washable AOC stage alongside True HEPA, 232 CADR over 360 sq ft. It scores 6.3, held back by a 28 dB floor and a 6 on quiet, and its PlasmaWave stage is switchable. Buy it for the carbon and use the switch. If a clinician or a specific allergy diagnosis has you set on the highest filter grade available, the Medify MA-40 is the H14-grade option at $290. Be clear-eyed about it: 5.0 overall, with 5s across clean air, quiet and smart, a 224 CADR that cheaper True HEPA units beat, a 31 dB floor that is loud for a bedroom, and manual controls. The UV stage is not what you are paying for.
Our picks for this
The decade-long reliability benchmark with eco mode
$170
7/10
True HEPA + washable carbon at a workhorse price
$160
6.3/10
H13/H14-grade filtration with UV at mid price
$290
5/10
FAQ
Is H13 or H14 filtration worth paying extra for?
For most homes, no. True HEPA already removes 99.97% of particles at the hardest size to catch, and H14's 99.995% is a difference you cannot perceive in a room. Air changes per hour matter far more than the grade. The catalog shows the cost of chasing it: the H14-grade Medify MA-40 is $290 for 224 CADR, while the True HEPA Coway AP-1512HH Mighty delivers 233 CADR for $170. Buy the grade only if a clinician specified it.
Should I turn PlasmaWave and other ionizers off?
In bedrooms and small closed rooms, yes. Ion emitters make particles settle onto surfaces rather than removing them from the home, and they can react with fragrance and cleaning-product chemicals to form formaldehyde and ultrafine particles. The filter does the actual work either way, so expect no meaningful loss in particle removal with the stage disabled. The Winix 5500-2's PlasmaWave is switchable, which is exactly the feature that makes it easy to live with.
Does UV-C in an air purifier actually kill germs?
Barely, in a consumer tower. Germicidal effect depends on dose, which is lamp intensity multiplied by exposure time, and air moving past a small internal lamp at these flow rates is exposed for a fraction of a second. The HEPA media beside it already captures bacteria, mold spores and droplet nuclei mechanically in one pass. The $90 GermGuardian AC4825E leads with UV-C and still scores 3.0 overall on 99 CADR and 153 sq ft of coverage.
Do air purifiers produce ozone?
Purely mechanical HEPA-plus-carbon units do not. Units with ion emitters or certain UV lamps can produce it as a by-product. Ozone is a lung irritant rather than a cleaner, and it reacts indoors to form formaldehyde and ultrafine particles. Before buying anything with an ionizer, check whether it appears on California's Air Resources Board certified list or carries UL 2998 validation for negligible ozone, and switch the stage off if you are unsure.
How often does the carbon filter need replacing?
Usually sooner than the HEPA, because carbon saturates. Once its pore surface is full, odors simply stop improving, and that is your replacement signal rather than a calendar date. Washable carbon like the Winix 5500-2's AOC stage is worth understanding properly: rinsing removes surface dust and restores airflow, but it does not regenerate carbon that has already adsorbed its fill, which needs industrial heat. Treat washing as maintenance and still plan on eventual replacement.
Can an ionizer replace a HEPA filter?
No. An ion emitter charges particles so they stick to walls, curtains and floors. They leave the air, and a sensor may read that as improvement, but they stay in the room and re-enter the air whenever you move around or vacuum. A filter physically removes and holds them. The only charged-particle design worth having is the kind that charges particles upstream of a filter that then collects them, as Blueair's HEPASilent does in the Blue Pure 211i Max.
What is the cheapest purifier here that is still worth buying?
The $100 Levoit Core 300 rather than the $90 GermGuardian AC4825E. Both are entry True HEPA towers with manual controls, but the Core 300 offers 141 CADR over 219 sq ft at a 24 dB floor and scores 5.0, while the AC4825E manages 99 CADR over 153 sq ft at 40 dB and scores 3.0. Ten dollars buys more clean air and a machine quiet enough to sleep beside. Neither has auto mode.
Will a purifier remove cooking smells and VOCs?
Partly. Activated carbon adsorbs many odor molecules and some volatile organic compounds, so a unit with a real carbon stage noticeably reduces cooking and pet smells. But plain carbon handles formaldehyde poorly, saturates over time, and does nothing at all for carbon dioxide, so a stuffy room stays stuffy after running one. For off-gassing from new furniture or fresh paint, opening a window still outperforms every purifier in this catalog.
See the full ranking
Every air purifier we track, scored on the specs that matter.
Best air purifiers of 2026