# Air purifiers

Buy an air purifier by its smoke CADR, the clean air it delivers per minute, sized to the room with the 2/3 rule. Then check two things listings blur: how loud it is at the speed your room needs, and whether its carbon is a light layer or a large filter. Ignore the stated coverage area.

## How it works

A portable air purifier is a fan that pushes room air through a filter. What it delivers is the clean air delivery rate (CADR): airflow multiplied by the share of particles the filter catches in one pass. A dense filter on a weak fan and a looser filter on a strong fan can earn the same CADR, so filter grade alone says little. AHAM measures CADR at top speed for smoke, dust and pollen; smoke, the smallest particles, gives the conservative figure. Room size follows from CADR: AHAM's label assumes 4.8 room volumes cleaned an hour. Gases and smells need a second stage, activated carbon, which works in proportion to how much there is and wears out as it fills. Electronic stages (ionizers, plasma, photocatalytic oxidation) charge or react with pollutants instead of trapping them, and some emit ozone or other byproducts. [5] [9] [2] [8] [11] [24] [10]

| Family | What it is | Suits |
|---|---|---|
| Fan and media filter | A pleated HEPA or near-HEPA filter, usually with a washable pre-filter and a carbon layer. CARB lists these as Mechanical. | Almost everyone; most of the catalog. |
| Filter plus ionizer | A media filter with an ion or plasma stage (PlasmaWave, bipolar device) that charges particles. CARB lists these as Electronic. | Buyers who do not mind an electronic stage, ideally one they can switch off. |
| Carbon-heavy gas unit | A unit built around kilograms of activated carbon, sometimes with modest particle CADR. | Smoke smell, cooking fumes, chemical odours, new-furniture smells. |
| Filterless electronic | Electrostatic precipitators and ionizers with plates or no filter at all; small CADRs and possible ozone. | Few buyers; reviewers who measure rarely recommend them. |
| Box-fan DIY (Corsi–Rosenthal box) | A box fan with MERV-13 furnace filters taped round it; cheap, loud, effective. Not a retail product but a useful price and performance reference. | Smoke emergencies and tight budgets. |

## How an expert narrows the choice

```mermaid
timeline
  section 1 · Smoke CADR
    Room under 200, 200 to 400, or over 400 square feet?
  section 2 · Gas and odour capacity
    heavy smells and fumes : ask Carbon mass
    everyday smells : ask Filter life
    particles only : skip Carbon mass
  section 3 · Ionizer or electronic stage
    Filter only, or is an ionizer fine?
```

1. **Smoke CADR.** Room under 200, 200 to 400, or over 400 square feet?
   The 2/3 rule turns the band's top edge into a smoke CADR floor: up to 133, up to 267, or at least 267 cfm. Ask exact size inside a band. Over 400 leaves about 31 percent of ENERGY STAR models.
   Skip if already known: room dimensions, an earlier room-size answer in this session.

2. **Gas and odour capacity.** Particles only, everyday smells, or heavy smells and fumes?
   CADR covers particles only. Each answer keeps units at that tier or above, so heavy smells and fumes leaves only the few units with a large carbon filter.
   If heavy smells and fumes: ask Carbon mass.
   If everyday smells: ask Filter life.
   If particles only: skip Carbon mass.
   Skip if already known: smoker in the home, stated problem is chemical fumes or VOCs, stated problem is pollen or dust only.

3. **Ionizer or electronic stage.** Filter only, or is an ionizer fine?
   Filter only keeps CARB's Mechanical units, about 60 percent of its certified listings. Ionizer fine adds the Electronic ones but still excludes ozone generators.
   Skip if already known: stated asthma or ozone concern, an earlier answer about ionizers.

## What to surface

### Gas and odour capacity

How much the unit can do about gases and smells: nothing (no carbon), everyday smells (a carbon layer or light carbon filter), or heavy smells and fumes (a separate carbon filter of 1.6 kg or more). CADR says nothing about smells or chemical vapours. Carbon removes many of them in proportion to how much there is, and light layers fill within months. HouseFresh calls the carbon in mainstream towers fine for mild household odours but not for serious gas problems; its recommended gas units all carry 1.6 kg or more. Carbon handles formaldehyde and ammonia poorly. [8] [9] [20] [24] [25] [22] [28]

Values: particles only, everyday smells, heavy smells and fumes.
Question to ask: Particles only, everyday smells, or heavy smells and fumes?

Used as: top-level split, filter. Data: needs research, backfill cost high. Look for it as: Listings and brand pages say 'activated carbon layer' or 'activated carbon filter' for both light and heavy stages, and almost never give a mass. Levoit calls the pelleted layer HouseFresh describes an 'Activated Carbon Layer'.
No carbon stated: particles only. Carbon stated without a mass: everyday smells. Heavy only when the brand or HouseFresh publishes 1.6 kg or more in a separate carbon filter. Mass needs a brand page or a tester, so backfill the top 100 and lazy-load the rest. The 1.6 kg line is one tester's; revisit it if a second source sets one.

### Smoke CADR

The clean air delivery rate for the smallest test particles: cubic feet of particle-free air delivered per minute at top speed. A 12 by 15 ft room (180 sq ft) needs at least 120. It is the one number that compares any two purifiers, and it decides which room a unit can serve. The 2/3 rule asks for a smoke CADR of at least two-thirds of the floor area in square feet at an 8 ft ceiling, so 200 sq ft needs about 130 cfm. Use smoke rather than dust or pollen: it is the smallest particle and the most conservative figure. Raise it for higher ceilings. [1] [3] [8] [9] [15] [16] [5] [27]

Range: Many HEPA units deliver about 150 to 300 cfm; the median ENERGY STAR model is 161 cfm.
Question to ask: Room under 200, 200 to 400, or over 400 square feet?

Used as: top-level split, filter, sortable. Data: needs research, backfill cost medium. Look for it as: Sometimes in a bullet as 'CADR 143 CFM for smoke'. Often missing even for AHAM-certified units. Brand spec blocks print an unlabelled 'CADR : 250 CFM'; treat it as smoke only when it matches the AHAM room size divided by 1.55. Values in m³/h divide by 1.699.
Brand page first, then the AHAM directory. The ENERGY STAR feed lists retail SKUs (LAP-C401S-WUSR), so match it only on an exact SKU. An AHAM or 4.8 ACH room size divided by 1.55 also works. With no CADR, keep the unit, mark it unknown and rank it below units that meet the floor; never use stated coverage.
Chart: Room-size scale.

### Ionizer or electronic stage

Whether the unit adds an electronic stage (ionizer, plasma, photocatalytic module) to its filter. 'Filter only' means mechanical only; 'ionizer fine' accepts any value except an ozone generator. Electronic stages can emit ozone, a lung irritant, and some form byproducts such as formaldehyde. California caps certified units at 50 ppb, but no standard measures the byproducts. Consumer Reports does not recommend ionizers, and in one tester's runs they added little to particle removal. Being able to switch it off matters: one popular model turns its ionizer back on after a power cut. [11] [12] [9] [8] [20] [13] [23]

Values: mechanical only, ionizer you can switch off, ionizer always on, intentional ozone generator.
Question to ask: Filter only, or is an ionizer fine?

Used as: top-level split, filter, shown on the card. Data: needs research, backfill cost low. Look for it as: Appears only as a brand name in the title or bullets: PlasmaWave, bipolar device, ionizer, plasma, PCO, UV.
Look the model up on CARB's certified list, which marks every unit Mechanical or Electronic. Switchability comes from the manual. Treat anything sold as an ozone generator as excluded.

### Dark display at night

Whether the display and air-quality lights can be switched off, by a button or a light sensor. Some units keep a coloured air-quality light or a display lit by default, which matters in a bedroom. Reviewers single out models that dim automatically in the dark, and brands list a lights-off option as a sleep feature. [22] [33] [32]

Question to ask: Need the lights fully off at night?

Used as: filter, shown on the card. Data: needs research, backfill cost low. Look for it as: Sometimes in a bullet ('You can turn off the light for sleeping', 'When the room is darkened ... Sleep Mode'); otherwise in the manual.
Manual or brand page lookup; only needed on the bedroom branch.

### Noise at rated CADR

How loud the unit is, in decibels, at the top speed where its CADR is measured. A purifier only works while it runs, and people turn down or switch off loud ones, especially in bedrooms. Brands print a range whose quiet end is the sleep setting, which delivers a fraction of the rated CADR; the loud end is the number that goes with the CADR. One tester found some brands understate their figures. About 50 dB is a running refrigerator. [9] [8] [20] [21] [27] [33] [31]


Used as: shown on the card, sortable. Data: needs research, backfill cost low. Look for it as: Brand spec blocks give a range such as 'Noise Level : 24–54.5dB'; take the top. Amazon's 'Noise Level' row shows the quiet end on some pages and the loud end on others, so do not trust it unlabelled.
One brand-page lookup per model. Record the top of the range and flag it as brand-stated; prefer a reviewer's measured figure where one exists.
Chart: CADR against noise.

### Quiet CADR

The clean air the unit delivers at its fastest speed under 45 dBA, as HouseFresh measures it: PM1 particles in a 728 cubic foot test room. The rated CADR is measured at top speed, and buyers rarely run it there overnight. EPA warns the CADR at lower speeds can be considerably lower, and HouseFresh finds the share kept at a quiet speed varies widely between models. It is the honest number for a bedroom. It is PM1, not AHAM smoke, so compare it with other quiet figures, not with rated CADR. [21] [22] [20] [9]


Used as: shown on the card, sortable. Data: needs research, backfill cost high. Look for it as: Never on listings or brand pages. Only HouseFresh publishes it in cfm ('Sub-45 dBA PM1 CADR'); Consumer Reports publishes low-speed ratings, not cfm.
Backfill from HouseFresh reviews only, top 100 models, and lazy-load the rest. Store the basis (PM1, HouseFresh room) with the value. Where no measurement exists, record unknown rather than estimate.
Chart: CADR against noise.

### Carbon mass

How much activated carbon the filter holds, in grams. EPA says carbon works when there is a large amount of it, and how much a filter holds decides how long it works before it fills. The range is wide: a tower purifier's carbon is a layer in its filter, while the Austin HealthMate carries up to 15 lb of carbon and mineral media. The same machine can take filters with different amounts. [8] [24] [25] [30]


Used as: shown on the card, sortable. Data: needs research, backfill cost medium. Look for it as: Rarely stated. Gas-focused brands such as Austin Air print it, sometimes for carbon mixed with other media; mainstream brands such as Levoit do not.
Brand page for gas-focused brands; HouseFresh roundups list it for many others, including per-filter figures for mainstream models. Limit to models with a carbon bed, top 100.
Chart: bar length on a log scale, so grams and kilograms fit on one axis.

### Filter grade

How efficient the main particle filter is in one pass, from true HEPA down to looser furnace-type media. True HEPA catches at least 99.97 percent of 0.3 µm particles (US definition) or 99.95 percent at the hardest size (European H13). 'HEPA-type' and 'HEPA-grade' have no fixed meaning. Grade matters less than buyers think: CADR already includes it, a denser filter can cut airflow, and a box fan with MERV-13 furnace filters cleans a test room about as fast as premium units. [6] [7] [9] [20] [5] [26] [18] [19] [16] [10]

Values: HEPA (99.97% at 0.3 µm, or H13 and up), HEPA-type or HEPA-grade, E-class or MERV media, no media filter.

Used as: shown on the card. Data: needs research, backfill cost medium. Look for it as: Amazon's 'Filter Type' row says 'HEPA' or a brand phrase such as '3 in 1 Filtration'. Neither proves the grade.
Map to HEPA only when the brand page names a test (IEST, EN 1822 or ISO 29463 class, or DOE 99.97 percent at 0.3 µm). Listing 'HEPA' or 'True HEPA' without a test maps to HEPA-type. Brand phrases with no grade map to unknown.

### CADR verification

Whether an independent lab has confirmed the CADR, through AHAM's Verifide programme, or the brand states it unchecked. AHAM Verifide models are picked and tested by an independent lab, so their CADR and label room size can be taken at face value; Consumer Reports trusts the seal and has found many unverified room-size claims too high. A brand-stated CADR is still useful but deserves a margin. [1] [3] [20] [32] [31] [13]

Values: AHAM Verifide, brand-stated CADR, no CADR stated.

Used as: shown on the card. Data: needs research, backfill cost low. Look for it as: Often in the title ('AHAM VERIFIDE') or Amazon's 'Specification Met' row; absent on many listings for certified models.
Check the AHAM directory by model number. ENERGY STAR certification is not proof: it requires an AHAM AC-1 test, not Verifide.

### Filter life

How many months the brand says each replaceable filter lasts before it should be changed. Filters are the running cost. Particle filters usually last 6 to 12 months. Light carbon layers fill faster: Consumer Reports says change carbon every three months, and HouseFresh found a Vital 200S's carbon ran out within six. Units with large carbon filters are rated far longer, 12 months to five years. Reviewers cost filters at the shortest stated interval. [20] [4] [21] [22] [24] [30] [28] [29]


Used as: shown on the card. Data: needs research, backfill cost low. Look for it as: On the replacement-filter page as a range ('6–12 months'); seldom on the purifier listing.
Store the shortest figure in the brand's range, separately for particle and carbon filters where they are sold apart.
Chart: Running cost bar.

## Numbers worth computing

| Metric | Formula | Why |
|---|---|---|
| Largest room it suits (sq ft) | `cadr_smoke * 1.5 * (8 / ceiling_height_ft), with ceiling_height_ft = 8 when unknown` | Turns CADR into the room size a buyer knows: the 2/3 rule run backwards. AHAM's label uses 1.55, which is 4.8 air changes an hour at 8 ft. The biggest certified unit suits about 860 sq ft; EPA suggests several units for larger spaces. [1] [8] [2] [16] [27] |
| Air changes per hour in your room (per hour) | `cadr_smoke * 60 / (room_sq_ft * ceiling_height_ft)` | The figure experts actually reason in. AHAM sizes for 4.8; below that, expect less than the 80 percent smoke reduction the rating assumes. [2] [9] [5] |
| Air changes per hour at a quiet speed (per hour) | `quiet_cadr * 60 / (room_sq_ft * ceiling_height_ft)` | What a bedroom gets overnight. HouseFresh measures PM1 in its own room, which includes that room's natural ventilation, so treat the result as a ranking against AHAM's 4.8 target, not a like-for-like count. [21] [20] [2] |
| Filter cost per year (USD per year) | `sum over filters of (filter_price * 12 / filter_life_months), using the shortest stated life` | The cost buyers forget. A cheap unit with short-lived proprietary filters can cost more over two years than a dearer one. Costing at the shortest stated life is HouseFresh's method. [21] [20] [28] [29] |
| Rated CADR per nameplate watt (cfm per W) | `cadr_smoke / nameplate power in watts` | A rough efficiency ranking between units, since purifiers should run all day. It is not comparable with the federal or ENERGY STAR figures, which use measured PM2.5 CADR per watt including standby. Nameplate overstates draw: Winix 5500-2 says 70 W, HouseFresh measured 53.46 W. [17] [13] [14] [23] [32] |
| Price per cfm of CADR (USD per cfm) | `price / cadr_smoke` | Reviewers rank value this way, since CADR is what the money buys. It shows how much of a premium buys features rather than clean air. [21] [23] |

## What you give up

- More clean air means more airflow, and at top speed that means noise. Consumer Reports' fix is a bigger unit run slower; AHAM warns that oversizing adds noise and energy use. [20] [9] Board branch: "See more units that stay quiet at your room size".
- Kilograms of carbon slow the air and take space. The Austin HealthMate's 15 lb of carbon and mineral media comes with a particle CADR that suits only a mid-sized room. [24] [30] [8] Board branch: "See more units built for smells and fumes".
- A denser filter catches more per pass but can slow the fan. Room cleaning depends on passes per hour, so a looser filter moving more air can win. [5] [26] [22] Board branch: "See more high-CADR units regardless of HEPA label".
- An ionizer can add a little particle removal on paper. It also adds a possible ozone source and byproducts no standard measures. [9] [20] [23] [11] Board branch: "See more filter-only units".
- Generic filters often cost half as much. The brand's CADR was measured with its own filter, so a generic one may fit loosely or filter less, and no one has tested it in that machine. [4] [31] [22] Board branch: "See more units with cheap genuine filters".

## How to draw it

**AHAM label** (bar, shows Smoke CADR, CADR verification). The seal on the box shows three CADR bars (smoke, dust, pollen) and one room size. Buyers who have shopped before have seen it, so drawing smoke CADR the same way needs no explanation. Used today by AHAM Verifide seal on certified packaging. [1] [9]

**Room-size scale** (scale, shows Smoke CADR). A horizontal cfm scale with the buyer's room requirement drawn as a line. Units left of the line are too small, which a table of numbers hides. Used today by EPA and ENERGY STAR publish the same sizing as a table; the scale is our proposal. [8] [15]

**CADR against noise** (two axis map, shows Smoke CADR, Noise at rated CADR, Quiet CADR). Each unit is a pair of points, top speed and quiet speed, on a CADR by decibel map, with the buyer's room requirement as a horizontal line. The bedroom pick is the unit whose quiet point clears the line. This is a proposal, nobody draws it this way yet. [21] [20]

**Running cost bar** (bar, shows Filter life). One stacked bar per unit: filters plus electricity per year. It makes a cheap unit with expensive filters visible next to a dearer, cheaper-to-run one. Used today by Consumer Reports and HouseFresh publish an annual running cost as a single figure; the bar is our proposal. [20] [21]

## Words you will meet

- **CADR (clean air delivery rate)**. The volume of particle-free air a purifier delivers per minute, in cubic feet per minute (cfm), measured at top speed.
- **AHAM**. The Association of Home Appliance Manufacturers, the US trade body that wrote the CADR test and runs the Verifide certification.
- **AHAM Verifide**. AHAM's seal for purifiers whose CADR an independent lab has confirmed.
- **2/3 rule**. AHAM's sizing rule: the smoke CADR should be at least two-thirds of the room's floor area in square feet.
- **ACH (air changes per hour)**. How many times an hour a room's full volume of air passes through the purifier.
- **Stated coverage**. The room size a brand prints, which is CADR divided by an air-change rate the brand chooses, so it differs widely between brands.
- **HEPA**. A filter that catches at least 99.97 percent of 0.3 µm particles in one pass, by the US definition.
- **H13**. The European filter class that catches at least 99.95 percent of particles at the size the filter finds hardest to catch.
- **HEPA-type**. A marketing label for a filter that does not meet the HEPA definition and has no fixed meaning.
- **MERV**. A 1 to 16 scale for furnace filters; MERV 13 is the grade used in box-fan DIY purifiers.
- **Activated carbon**. Charcoal processed to have a huge surface area, which holds gas and odour molecules until it fills up.
- **VOCs**. Volatile organic compounds: gases given off by paints, cleaning products, cooking and new furniture.
- **Ionizer**. An electronic stage that charges particles so they stick to the filter or to room surfaces; sold under names like PlasmaWave or bipolar device.
- **Ozone**. A reactive gas that irritates the lungs, emitted by some electronic air cleaners.
- **CARB certified**. Approved by the California Air Resources Board, which every air cleaner sold in California must be; it limits ozone but does not judge effectiveness.
- **ppb**. Parts per billion, the unit for ozone limits; CARB's ceiling for certified air cleaners is 50.
- **dB**. Decibels, the unit of loudness; about 50 dB is a running refrigerator.
- **Corsi–Rosenthal box**. A home-made purifier: a box fan with four or five MERV-13 furnace filters taped around it.
- **IEF (integrated energy factor)**. The federal and ENERGY STAR efficiency measure: CADR per watt, counting standby power.

## Where we made a call

- Stated room coverage is marketing; rank and filter on smoke CADR instead. (confidence high). Levoit prints coverage at two air-change rates 4.8 times apart; Coway quotes 361 to 1,800 sq ft for one family; Consumer Reports found many claims too high. Ours, built on those sources. [27] [29] [33] [20] [2]
- A HEPA label matters less than CADR; do not filter on it. (confidence medium). Sources disagree. Consumer Reports and EPA lead with HEPA. HouseFresh and the box-fan results show MERV-13 cleaning rooms fast, and CADR already includes filter efficiency. We side with CADR because it measures the room outcome. [20] [8] [26] [22] [5]
- Prefer filter-only units; accept an ionizer only if it switches off. (confidence medium). Sources disagree. CARB certifies ionizers at up to 50 ppb ozone, and ENERGY STAR test data show certified ones far below that. Consumer Reports does not recommend them, EPA notes unmeasured byproducts, and HouseFresh found little performance gain. We side with caution. [11] [16] [20] [9] [23]
- For heavy smells and fumes, carbon mass is the signal; a carbon layer in the main filter handles only everyday smells. (confidence medium). EPA, Consumer Reports and HouseFresh agree that light carbon saturates quickly and more carbon works longer. The 1.6 kg line for 'heavy' is HouseFresh's recommended list, one source. [8] [9] [20] [24]
- Activated carbon does little for formaldehyde. (confidence high). EPA's technical summary and Consumer Reports both name formaldehyde and ammonia as gases carbon handles poorly. [9] [20]
- For a bedroom, buy more CADR than the room needs and run it on a low speed. (confidence medium). Sources disagree. Consumer Reports advises sizing up to run slow, and EPA says loud units get switched off. AHAM warns an oversized unit brings unnecessary energy use and noise. We side with Consumer Reports for bedrooms, since overnight use is at low speed. [20] [9] [1] [3]
- Brand noise figures are a weak guide; the listing's single noise number is often the sleep setting. (confidence medium). Amazon shows the quiet end on two of three pages read and the loud end on the third. HouseFresh reports some brands understate noise. EPA says noise is seldom reported in a standard way. [31] [32] [33] [21] [9]
- Generic replacement filters are a reasonable saving, with the caveat that the rated CADR no longer applies. (confidence low). Sources disagree. AHAM, the manufacturers' body, warns against them. HouseFresh points buyers to them at half price while noting performance varies. Levoit's own disclaimer limits its CADR to its own filters. No independent test settles it. [4] [22] [31]

## References

1. [AHAM Verifide: How to choose the best air cleaner for your home](https://ahamverifide.org/how-to-choose-the-best-air-cleaner-for-your-home/) (standards body). Supports: room size is the first thing to know; 2/3 rule: smoke CADR at least two-thirds of floor area; 200 sq ft needs 132; corners reduce performance.
2. [AHAM Verifide: Air Changes per Hour FAQ](https://ahamverifide.org/ach-faq/) (standards body). Supports: AHAM's suggested room size is based on 4.8 air changes per hour.
3. [AHAM Consumer Blog: Tips for Choosing the Ideal Portable Room Air Cleaner (Dec 2024)](https://blog.aham.org/tips-for-choosing-the-ideal-portable-room-air-cleaner/) (standards body). Supports: the two-thirds rule restated; AHAM considers CADR the most important comparison metric; Verifide units are independently lab-tested.
4. [AHAM Consumer Blog: Why Genuine Replacement Air Filters Matter (Oct 2025)](https://blog.aham.org/why-genuine-replacement-air-filters-matter-for-air-cleaners/) (standards body). Supports: filters typically replaced every 6 to 12 months; non-genuine filters may cost half and fit poorly.
5. [Wikipedia: Clean air delivery rate](https://en.wikipedia.org/wiki/Clean_air_delivery_rate) (primer). Supports: CADR is airflow times single-pass efficiency; denser filters can lower airflow and CADR; 2/3 rule assumes 80 percent smoke removal.
6. [Wikipedia: HEPA](https://en.wikipedia.org/wiki/HEPA) (primer). Supports: HEPA is 99.97 percent at 0.3 µm (US) or 99.95 percent (EN, ISO); True HEPA has no legal meaning; HEPA-type does not meet HEPA.
7. [EN 1822: HEPA and ULPA filter classes](https://www.en1822.info/) (primer). Supports: H13 means 99.95 percent efficiency at the most penetrating particle size.
8. [US EPA: Guide to Air Cleaners in the Home](https://www.epa.gov/indoor-air-quality-iaq/guide-air-cleaners-home) (regulator). Supports: minimum CADR 65 cfm per 100 sq ft at an 8 ft ceiling; CADR covers particles only; carbon works when there is a lot of it; avoid ozone-producing cleaners; about 50 dB is a modern refrigerator; CADR is measured at the highest speed.
9. [US EPA: Residential Air Cleaners, A Technical Summary, 3rd edition (2018)](https://www.epa.gov/sites/default/files/2018-07/documents/residential_air_cleaners_-_a_technical_summary_3rd_edition.pdf) (regulator). Supports: ratings are at maximum airflow and noise; lower speeds may give much lower CADR; noisy units get turned off or stored; smoke CADR is the conservative figure; many HEPA units deliver about 150 to 300 cfm; thin carbon layers saturate quickly; carbon handles formaldehyde poorly; plasma, ionizers and PCO can emit ozone or byproducts.
10. [US EPA: Wildfires and Indoor Air Quality](https://www.epa.gov/emergencies-iaq/wildfires-and-indoor-air-quality-iaq) (regulator). Supports: box-fan DIY cleaners are a fallback; units with known CADR preferred.
11. [California Air Resources Board: California Certified Air Cleaning Devices](https://ww2.arb.ca.gov/our-work/programs/air-cleaners-ozone-products/california-certified-air-cleaning-devices) (regulator). Supports: certification required since 2010; electronic units capped at 50 ppb ozone; each model listed Mechanical (about 60 percent of rows) or Electronic; certification does not judge effectiveness.
12. [California Air Resources Board: Air cleaner information for consumers](https://ww2.arb.ca.gov/our-work/programs/air-cleaners-ozone-products/air-cleaner-information-consumers) (regulator). Supports: all air cleaners sold in California, including online, must be CARB certified.
13. [ENERGY STAR: Room Air Cleaners Key Product Criteria](https://www.energystar.gov/products/air_purifiers_cleaners/key_product_criteria) (standards body). Supports: minimum IEF 4.4, 5.4 and 5.6 by PM2.5 CADR bin; ozone at most 50 ppb for certification.
14. [ENERGY STAR Version 3.0 Room Air Cleaners Final Draft Specification and Final Cover Memo (Jan 2025)](https://www.energystar.gov/sites/default/files/2024-11/ENERGY%20STAR%20Version%203.0%20Room%20Air%20Cleaners%20Final%20Draft%20Specification.pdf) (standards body). Supports: minimum IEF 4.4, 5.4 and 5.6; effective 9 October 2025 with no changes after the final draft.
15. [ENERGY STAR: Room air cleaners](https://www.energystar.gov/products/air_purifiers_cleaners) (standards body). Supports: same room-size to minimum-CADR table as EPA.
16. [ENERGY STAR Certified Room Air Cleaners V3.0 (data.energystar.gov; 247 rows, about 190 model numbers)](https://data.energystar.gov/resource/gaa3-swy6.json) (standards body). Supports: label room size equals 1.55 times smoke CADR on every row; median smoke CADR 161 cfm; 31 percent at 267 or more; largest 577; 218 of 247 rows self-report HEPA-Type filters; certified ionizers measure far below the 50 ppb cap.
17. [10 CFR 430.32(ee), energy conservation standards for air cleaners (Cornell LII)](https://www.law.cornell.edu/cfr/text/10/430.32) (regulator). Supports: from 31 Dec 2025, minimum IEF 1.9, 2.4 and 2.9 CADR per watt by size.
18. [BBB National Programs NAD: Vesync voluntarily discontinues certain HEPA claims (Aug 2023)](https://bbbprograms.org/media/newsroom/decisions/vesync-hepa-air-purifiers) (regulator). Supports: Levoit dropped HEPA claims for Core 300 and EverestAir after a Dyson challenge; merits not reviewed.
19. [Open Class Actions: Levoit air purifier HEPA settlement](https://openclassactions.com/settlements/false-advertising/levoit-air-purifier-hepa-class-action-settlement.php) (other). Supports: settlement covers Levoit Core and EverestAir HEPA, True HEPA and H13 labels, 2019 to 2023; Vesync denies wrongdoing.
20. [Consumer Reports: Air purifier buying guide](https://www.consumerreports.org/appliances/air-purifiers/buying-guide/) (specialist review). Supports: tests CADR at top speed and at 50 dB or less; size up and run low for quiet; carbon replaced about every three months, particle filters 6 to 12; does not recommend ionizers; AHAM seal trustworthy; many brand room claims too high.
21. [HouseFresh: How we test air purifiers](https://housefresh.com/how-we-test-air-purifiers/) (specialist review). Supports: scores CADR at top speed and the highest speed under 45 dBA; some brands understate noise; costs filters at the shortest stated life; value scored as cost per CADR.
22. [HouseFresh: Levoit Vital 200S review](https://housefresh.com/levoit-vital-200s-review/) (specialist review). Supports: measured CADR at top and quiet speeds, noise and power per speed; bonded filter's pelleted carbon usually runs out within six months; generic filters at half price, performance varies; light-detection mode for bedrooms.
23. [HouseFresh: Winix 5500-2 review](https://housefresh.com/winix-5500-2-review/) (specialist review). Supports: ionizer on or off made one minute's difference in the room test; ionizer turns back on after a power cut; measured 53.46 W at top speed against 70 W nameplate.
24. [HouseFresh: The best activated charcoal air purifiers](https://housefresh.com/air-purifier-technologies-activated-carbon-air-filter/) (specialist review). Supports: mainstream units' carbon fine for mild odours, not serious gas problems; recommended gas units carry 1.6 kg or more; carbon-bed filters rated 12 months to five years; Austin HealthMate: 15 lb carbon and zeolite, modest particle CADR.
25. [HouseFresh: The best air purifiers for VOCs, odors and gases](https://housefresh.com/air-purifier-for-vocs/) (specialist review). Supports: cheap units use carbon-impregnated fabric that is not enough for most VOC problems.
26. [HouseFresh: Corsi–Rosenthal Box review](https://housefresh.com/corsi-rosenthal-box-review/) (specialist review). Supports: a MERV-13 box fan cleared PM1 as fast as some $900 units; HEPA grade is not required for clean air.
27. [Levoit: Vital 200S-P product page](https://levoit.com/products/vital-200s-smart-true-hepa-air-purifier) (spec sheet). Supports: ideal room size 1,862 sq ft at 1 ACH or 388 at 4.8 ACH; CADR 250 cfm, rated power 50 W, noise 23 to 54 dB.
28. [Levoit: replacement filter pages (store product feed)](https://levoit.com/products.json) (spec sheet). Supports: Vital 200S-P filter $49.99, lasts 6 to 12 months; carbon described as a layer; Core 300 filter replacement every 6 to 8 months.
29. [Coway: Airmega product and filter pages (store product feed)](https://cowaymega.com/products.json) (spec sheet). Supports: Mighty family coverage stated as 361, 788, 1,680 and 1,800 sq ft; HEPA one year, carbon six months; AP-1512HH has a bipolar device.
30. [Austin Air: HealthMate product page](https://austinair.com/products/healthmate/) (spec sheet). Supports: up to 15 lb of activated carbon and mineral media; filter rated five years.
31. [Amazon listing: Levoit Core 300-P (B07VVK39F7)](https://www.amazon.com/dp/B07VVK39F7) (retail listing). Supports: Floor Area 1,073 sq ft is the 1 ACH figure; Noise Level 24 dB is the sleep setting; CADR applies only to Levoit's own filters.
32. [Amazon listing: Winix 5500-2 (B01D8DAYII)](https://www.amazon.com/dp/B01D8DAYII) (retail listing). Supports: no CADR number shown; Floor Area 360 is the AHAM size; Noise Level 27.8 dB is the quiet end; auto sleep mode when the room is dark.
33. [Amazon listing: Coway AP-1512HH (B01728NLRG)](https://www.amazon.com/Coway-AP-1512HH-Mighty-Purifier-White/dp/B01728NLRG) (retail listing). Supports: title says 1,748 sq ft, table says 361; Noise Level 53.8 dB is the loud end; the light can be turned off for sleeping.