How we rank products

Every ranking on this site is calculated, not hand-picked. Here is the whole method.

1. Specs come from the manufacturer

We enter only figures the manufacturer publishes and link the page we used under every product. If a figure isn't published we show "Not listed" instead of guessing.

2. Each spec is scaled from 0 to 1

Each criterion has a fixed reference range, from a "weakest reasonable" value to a "best in class" value (for example 200 to 500 CFM of smoke CADR for large-room purifiers). A product's figure is placed on that range and clamped between 0 and 1, so scores stay meaningful as products are added. For specs where lower is better, such as noise in decibels, the range runs the other way. Yes/no features count as 1 or 0. The ranges are listed below.

3. Weights decide what matters most

Each category lists its criteria and weights on the page. The score is the weighted average, multiplied by 10:

score = 10 × Σ(weight × scaled value) ÷ Σ(weight)

4. How missing data is scored

If a headline spec isn't published (CADR, CADR per watt, sleep-mode noise), the product scores at the lowest value any product in that category actually published, so an unpublished figure never outranks a published one. The page marks each such cell with an asterisk. A few criteria that only apply to some products (like top-speed noise or filter life, which some makers don't publish) are skipped instead, and their weight is removed for that product.

5. Prices and ratings switch on automatically

When our retailer feeds supply a current price or customer rating for at least 60% of products in a category, two more criteria join the formula: Price / value (20%) and Customer rating (15%, discounted when there are few reviews). Until then we show "Check price" and never estimate a price.

6. What the score can't tell you

Specs don't capture build quality, sensor accuracy, real-world odor removal or support. CADR is the maker's own figure (often AHAM-verified); noise is measured at different distances by different makers; filter life assumes each maker's own daily run time. Use the score to shortlist, then read the pros, cons and sources. When we add hands-on testing notes, we label them clearly.

7. Money never changes the order

We earn affiliate commissions from some retailers. Commission rates are not an input to the formula, and no brand can pay for placement. See our affiliate disclosure.

Criteria by category

Large-Room Air Purifiers

• Smoke CADR: 40% · range 200 → 500 CFM
• CADR per watt: 15% · range 3 → 10 CFM/W
• Quietest setting: 10% · range 30 → 17 dB
• Loudest setting: 10% · range 60 → 46 dB · skipped when not published
• Carbon or odor stage included: 10% · yes/no
• Main filter life: 15% · range 6 → 24 months · skipped when not published

Bedroom & Small-Room Air Purifiers

• Quietest setting: 30% · range 32 → 17 dB
• Smoke CADR: 25% · range 50 → 250 CFM
• Loudest setting: 15% · range 56 → 42 dB · skipped when not published
• CADR per watt: 10% · range 3 → 10 CFM/W
• Carbon or odor stage included: 10% · yes/no
• Main filter life: 10% · range 6 → 24 months · skipped when not published

HEPA + Carbon Purifiers for Smoke & Odor

• Smoke CADR: 40% · range 150 → 450 CFM
• Carbon/odor filter life: 20% · range 3 → 24 months · skipped when not published
• Smoke or VOC filter offered: 15% · yes/no
• CADR per watt: 15% · range 3 → 10 CFM/W
• Loudest setting: 10% · range 60 → 46 dB · skipped when not published

Purifier + Fan, Heater & Humidifier Combos

• Heater mode: 20% · yes/no
• Cooling fan / circulator mode: 15% · yes/no
• Humidifier: 15% · yes/no
• Smoke CADR: 25% · range 100 → 250 CFM
• Oscillation range: 10% · range 45 → 350 ° · skipped when not published
• Carbon stage included: 15% · yes/no