The formula
score = 5 × (0.25·C1 + 0.15·C3 + 0.25·C4 + 0.15·C5 + 0.2·C6), rounded to one decimal. Each component is a number from 0 to 1; the weights sum to 1. Below is what each one measures, exactly how it's calculated, and why it carries the weight it does.
| Component | Formula | Weight | Why this weight |
|---|---|---|---|
| C1 - CADR independently verified? | 1.0 if AHAM Verifide-listed, else 0.5 | 25% | CADR is the core spec of the whole product category - how much air the unit actually cleans. An unverified number is a real, unresolved information gap on the single most load-bearing spec, not a minor footnote. |
| C3 - CARB status | 1.0 if Mechanical, 0.85 if Electronic | 15% | Both categories pass the same mandatory ozone-safety certification - see our full breakdown. The small, deliberately non-punitive gap reflects that Mechanical is the more conservative default we recommend, not that Electronic fails anything. |
| C4 - real annual filter cost | 1 − (cost − $40) ÷ ($220 − $40), clamped to [0, 1] | 25% | Direct cost to whoever buys the unit, and the second-largest finding of the audit that built this site's data (see true filter cost) - PuroAir runs 2-3× more expensive than it looked before we checked. $40 and $220 are fixed dollar anchors, not the current cheapest/priciest model - see "why fixed anchors" below. |
| C5 - filtration class | 1.0 if standard True HEPA H13 (99.97% @ 0.3μm), 0.8 if a different mechanism or threshold | 15% | The smallest weight of the five, deliberately - there's no independent registry for HEPA classification the way AHAM covers CADR or CARB covers ozone. This is drawn from each manufacturer's own technical claims, checked for consistency, but not third-party verified the same way. |
| C6 - CADR relative to a typical room | min(smoke CADR ÷ 250, 1.0) | 20% | Without this, a small, weak unit that's merely cheap to run scores artificially well - C4 rewards low cost regardless of whether the unit can actually clear a normal-sized room. 250 CFM is roughly what a typical living room needs under the 2/3 rule; it's our own reference point, not an industry standard - see "why 250" below. |
What score does NOT measure
This matters more than it looks like it should, so it gets its own section rather than a footnote:
- Whether the marketing was honest. That's a separate flag (see below), shown next to score, never folded into it. A product can score well and still carry a "spec overstated" warning - those are two different questions.
- How well a specific unit performs in your specific room. C6 checks CADR against a fixed 250 CFM reference, not against the room you actually have. A 34 CFM unit scoring well on every other axis is still the wrong purifier for a 400 sq ft living room - check the 2/3-rule math on the model's own review before buying regardless of score.
- Durability, noise character, build quality, customer support, or design. We haven't independently tested any of these across the catalog the way we've checked CADR, CARB, and filter cost - they show up as qualitative notes in individual reviews, not as score inputs.
The Core Mini paradox
Levoit Core Mini-P scores 4.1/5 - one of the highest scores on the site - while carrying this site's largest "spec overstated" flag: its manufacturer claims 65 CFM; AHAM Verifide shows 34 (91% overstated). Here's why that's not a contradiction.
Score evaluates the product. The flag evaluates how it was sold. 34 CFM is an honest, adequate number for the room this unit is actually recommended for - a closet, a small bathroom, a litter box alcove under about 50 sq ft. On every other axis, it's a genuinely strong unit: AHAM-verified (so C1 is full marks), Mechanical (full marks on C3), the cheapest filter to run in the catalog (near-full marks on C4), and standard True HEPA (full marks on C5). It only loses ground on C6, because 34 CFM is a small fraction of the 250 CFM reference point - appropriately, since it's a small-room unit being measured against a living-room yardstick.
The problem was never the device. The problem was a number on the box that didn't match reality. Fixing that number (and flagging that it needed fixing) is a separate job from judging whether the device, once correctly understood, is worth buying for the room it's actually suited to. It is.
Honest limitations
- C5 has no independent registry. Unlike AHAM (CADR) or CARB (ozone), nothing certifies "True HEPA" the way those two certify their own specs. We classify based on each manufacturer's own published grade and threshold claims, checked for internal consistency, not against a third party.
- C6's 250 CFM reference is our choice, not a standard. It's a reasonable approximation of what a typical living room needs under the 2/3 rule (roughly 375 sq ft of coverage), not a figure any standards body publishes. A different, equally defensible reference point would shift every model's C6 by some amount - we picked one, disclosed it, and kept it fixed rather than adjusting it per model.
- The $40-$220 cost anchors are fixed, not relative to the current 9-model set. This is deliberate: if we normalized against today's cheapest and priciest model, adding a 10th model with a new extreme price would silently shift every existing score, even for products whose own data never changed.
All 9 models, computed
| Model | C1 | C3 | C4 | C5 | C6 | Score | Flag |
|---|---|---|---|---|---|---|---|
| Levoit Vital 200S-P | 1.00 | 1.00 | 0.67 | 1.00 | 1.00 | 4.6 | - |
| Winix 5520 | 1.00 | 0.85 | 0.78 | 1.00 | 1.00 | 4.6 | - |
| Levoit Core 300-P | 1.00 | 1.00 | 0.89 | 1.00 | 0.57 | 4.4 | ⚠ 36% overstated |
| Blueair Blue Pure 211i Max | 1.00 | 0.85 | 0.71 | 0.80 | 1.00 | 4.4 | - |
| Levoit Vital 100S-P | 1.00 | 1.00 | 0.78 | 1.00 | 0.57 | 4.3 | ⚠ 36% overstated |
| Levoit Core Mini-P | 1.00 | 1.00 | 0.94 | 1.00 | 0.14 | 4.1 | ⚠ 91% overstated |
| uahpet Air Purifier Slim | 0.50 | 1.00 | 0.71 | 1.00 | 0.48 | 3.5 | - |
| AirDoctor AD3500 | 0.50 | 0.85 | 0.27 | 1.00 | 1.00 | 3.3 | - |
| PuroAir 240 | 0.50 | 1.00 | 0.02 | 0.80 | 0.82 | 2.8 | - |
Recompute any row yourself: C1/C3/C4/C5/C6 use only AHAM/CARB status and filter cost, both published with sources. Plug them into the formula above - you should land on the same number shown here.
Adding a new model later
When a 10th model joins the site, it gets scored against the same fixed anchors - $40-$220 for cost, 250 CFM for the CADR reference - not against whatever the other 9 happen to look like at the time. That means existing scores don't move just because a new model was added. A score only changes when that specific product's own underlying data changes: a corrected CADR figure, a new CARB filing, or a manufacturer's filter price shifting.
This page and the formula it runs were last verified against registry check and true filter cost data on 2026-08-13.