How AI Search Is Recommending Air Purifiers: Monthly Trends

Executive Summary

Levoit remains the category leader in AI search recommendations for air purifiers, holding 86.8% valid recommendation coverage in October 2026, a stable position at the top. The gap to the next brand, Coway at 80.1%, stands at 6.7 points — a notably narrower gap than earlier in the series, reflecting Coway's three-month climb rather than any Levoit decline.

Coway is the largest riser across the full series, up 9.9 points from 70.2% in July 2026 to 80.1% in October 2026, a significant move that has now continued for three consecutive months. It also posted the month's sharpest single-month move, up 7.0 points from September's 73.1%. In absolute terms, Coway recorded 490 valid recommendations in October versus 432 in July. Blueair also moved up from its prior month, gaining 5.6 points from September to October, a significant one-month shift.

Dyson is the largest decliner across the full baseline-to-current series, down 12.7 points from 24.6% in July 2026 to 11.9% in October 2026, a significant decline, though October itself was effectively flat against September. Austin Air fell 8.0 points from 23.4% in July to 15.4% in October, a significant decline that has now continued for three months. Rabbit Air declined 5.2 points across the same window, also significant.

Against the prior month, the category was mixed rather than uniform: Coway, Blueair, and Winix gained, while IQAir posted the sharpest single-month decline, down 6.1 points from September's 45.8%, a significant move on the month. The trend series for this vertical now spans July 2026 through October 2026, with four comparable monthly runs.

Each monthly run begins with 800 prompt-surface observations across the benchmark's defined AI/search surface universe. In July 2026, this yielded 504 unique questions; 800 mentioned a tracked brand or competitor; 787 were relevant and 13 were irrelevant, producing 615 qualified observations. In October 2026, the run yielded 543 unique questions; 800 mentioned a tracked brand or competitor; 774 were relevant and 26 were irrelevant, producing 612 qualified observations. Intermediate months contributed 700 collected prompts (497 unique questions, 546 qualified observations) in August 2026 and 800 collected prompts (545 unique questions, 598 qualified observations) in September 2026. The public metrics use the 612 qualified October observations that survive both qualification stages.

Key Findings

AI Response Inconsistency Alerts

Five critical or high-severity factual inconsistencies were detected in this month's run, across four AI platforms: ChatGPT, Copilot, Gemini, and Perplexity.

Blueair

Three high-severity conflicts were identified for Blueair, all concerning the room coverage specification of the Blue Pure 311i Max. All three arose from the same question, "What is a good hepa air purifier?"

ChatGPT stated the Blue Pure 311i Max is suited to large rooms and open spaces, citing a Best Buy listing describing it as rated for 1,800 sq ft, referencing RTINGS.com and TechGearLab. Gemini stated the same purifier is for medium rooms at approximately 380 sq ft. These two coverage claims are incompatible for a single product.

A second conflict saw ChatGPT's 1,800 sq ft claim set against Copilot's stated coverage of 929 sq ft. Copilot cited a Walmart product listing, an IQAir product page, and a Filtrete filter listing. A flagged source, an xcnnews.com article, was cited on the Copilot side describing the unit as covering 929 sq ft.

Across the three conflicts, AI platforms gave three different coverage figures for the same purifier: 1,800 sq ft, 929 sq ft, and approximately 380 sq ft.

Coway

One high-severity conflict was identified for Coway, concerning the room coverage of the Airmega 400. When asked "What is a good hepa air purifier?", Copilot stated the Airmega 400 cleans up to 1,560 sq ft, while Perplexity stated the unit is for medium to small rooms. A purifier rated for 1,560 sq ft cannot simultaneously be a medium-to-small-room unit.

Copilot cited a Walmart product listing, an IQAir product page, and a Filtrete filter listing. Perplexity cited an Amazon Best Sellers page, Consumer Reports, and The New York Times Wirecutter. A flagged source, the xcnnews.com article, was cited on the Copilot side describing the Airmega 400 as suited to large rooms at 1,560 sq ft.

Molekule

One high-severity conflict was identified for Molekule, concerning whether the company remains in business. When asked "Is a Molekule worth it?", Copilot stated that Molekule faced bankruptcy in 2023, though it continues to operate with updated models. ChatGPT presented Molekule as an active, currently operating company, with no mention of bankruptcy. These two accounts cannot both be true.

Copilot cited a US Air Purifiers product page and two Amazon product listings. ChatGPT cited Molekule's own site and a Tom's Guide review. Two flagged sources were cited on the ChatGPT side: a Consumer's Best review describing Molekule as a thoughtful choice for style-conscious homes, and Molekule's own site describing its ranges as premium PECO HEPA air purifiers. A third flagged source was cited on the Copilot side, describing a Chapter 11 bankruptcy filing in 2023.

Benchmark Context

The report separates the raw collection universe from the qualified analysis set. Brand-level recommendation percentages are calculated within the qualified benchmark set, which is smaller than the raw collection due to relevance filtering and reserved prompts.

Both July and October recorded qualified observations across all six AI surface families. August dipped to five families before September restored breadth to six, which affects the comparability of absolute counts in that intermediate month.

Benchmark-Level Metrics

The recommendation-shaped answer share strengthened notably across the series, rising from 63.7% in July to 70.8% in October, indicating a larger proportion of qualified answers took the form of a recommendation shortlist as the series progressed.

AI Recommendation Trend

Questions This Section Answers

• Which air purifier brands lead AI search recommendations, and how has the top tier tightened?
• How large is the gap between Levoit and Coway, and which brands gained or lost coverage across the series?

Levoit holds the lead as Coway closes the gap and the top tier tightens

Levoit's leadership is intact across all four months, but the distance to Coway has narrowed steadily. Coway's three-month climb has brought it within 6.7 points of the leader in October, the closest the two brands have been across the series.

The single largest baseline-to-current shifts — Coway's 9.9-point rise and Dyson's 12.7-point decline — are classified as significant. The remaining movement across the category reflects the combination of several smaller, mostly non-significant changes rather than one dominant swing, with the exception of Winix's 9.6-point gain in top-three placement and Blueair's 11.8-point gain in the same measure, both significant.

What Changed This Month

Questions This Section Answers

• What drove Coway's largest single-month coverage gain?
• What explains Dyson's and Austin Air's continued declines in AI recommendations?
• Did IQAir's October pullback reverse its earlier climb?

Coway: Significant three-month riser closing on the lead

Coway's valid recommendation coverage rose 9.9 points from 70.2% in July 2026 to 80.1% in October 2026, a significant move that has now continued for three consecutive months. The most recent step was the sharpest: coverage rose 7.0 points from September's 73.1%, the largest single-month move in the category.

In absolute terms, Coway recorded 490 valid recommendations in October versus 432 in July. Its top-three placement rose 17.0 points across the series, from 52.8% to 69.8%, and its rank-one rate rose 15.1 points, from 19.4% to 34.5% — both significant.

The distinction worth noting is that Coway's gains span presence, top-three placement, and rank-one placement together, rather than concentrating in a single measure. Its raw mention presence rate rose 5.9 points to 84.6% over the series.

Highest-priority diagnostic: Which prompt types or surfaces are driving the rank-one gains, and is the rise concentrated in a few high-visibility contexts or spread across the full question set?

Dyson: Largest full-series decliner, now flat month to month

Dyson's valid recommendation coverage fell 12.7 points from 24.6% in July 2026 to 11.9% in October 2026, a significant decline that remains the largest in the category. October itself was effectively flat against September, up 0.4 points from 11.5%, so the bulk of the series decline occurred in the earlier months.

In absolute terms, Dyson recorded 73 valid recommendations in October versus 151 in July. Its top-three rate fell 4.5 points, from 7.6% to 3.1%, a significant decline, and its rank-one rate slipped from 1.0% to 0.5%.

The distinction to notice is that Dyson's raw mention presence rate fell 16.3 points to 18.3% across the series — a steeper decline than its coverage — meaning the brand is appearing in fewer answers overall even as its recommendation share stabilizes.

Highest-priority diagnostic: Which prompts stopped surfacing Dyson entirely earlier in the series, and which brand now appears in those contexts instead?

Austin Air: Significant three-month decline in presence and coverage

Austin Air's valid recommendation coverage fell 8.0 points from 23.4% in July 2026 to 15.4% in October 2026, a significant decline that has now continued for three months. The sharpest step came this month, down 5.0 points from September's 20.4%.

In absolute terms, Austin Air recorded 94 valid recommendations in October versus 144 in July. Its raw mention presence rate fell 8.7 points to 17.8% across the series, a significant decline.

The distinction worth noting is that Austin Air's rank-one rate actually rose 1.8 points to 4.7% over the series, though the change is not significant. The brand is appearing in fewer answers overall while its top-position rate has edged up on a small base.

Highest-priority diagnostic: Which prompts stopped surfacing Austin Air across the three-month decline, and is the pattern concentrated on specific surfaces?

IQAir: Sharpest single-month decline after a mid-series climb

IQAir's valid recommendation coverage fell 6.1 points from September's 45.8% to 39.7% in October 2026, a significant move on the month. Across the full series, coverage still rose 4.7 points from 35.0% in July, a non-significant change, so the October decline partially reversed an earlier climb.

In absolute terms, IQAir recorded 243 valid recommendations in October versus 215 in July. Its raw mention presence rate rose 3.1 points to 43.8% across the series.

The distinction to notice is the divergence between the series and the month: across four months IQAir is up, but the most recent month reversed part of that gain. This is the first decline in IQAir's run and breaks a two-month upward pattern.

Highest-priority diagnostic: Whether September's higher coverage reflected a temporary surge or whether October's pullback reflects a durable shift, and which prompts drove the change.

The widening Coway to Dyson gap

The benchmark flags a notably widening gap between Coway and Dyson: the two brands were 45.6 points apart in July 2026 and 68.2 points apart in October 2026, widening in every month of the series. This is a category-level observation about competitive repositioning at the top and bottom of the recommendation range, not a claim about either brand's underlying product position.

A second widening gap runs between Austin Air and Coway, from 46.8 points in July to 64.7 points in October, also widening every month.

Highest-priority diagnostic: Whether the divergence reflects a small number of influential evidence sources shifting, or a broad repositioning across many prompts.

Buyer-Intent Interpretation

Questions This Section Answers

• What buyer-intent cluster do AI air purifier recommendations actually capture?
• What commercial questions can the benchmark not yet answer about pricing or head-to-head comparisons?

In October 2026, all 612 qualified observations fell into the Brand Recommendation cluster. No observations qualified for the pricing and value or multi-brand comparison clusters, matching the July, August, and September profiles.

The public benchmark therefore answers the question of which brands AI systems recommend for air purifier needs. It cannot yet speak to how brands are positioned on price or value, nor which brand wins in explicit head-to-head comparisons. Those commercial questions require additional research beyond the current benchmark scope.

Brand Opportunity Summary

Questions This Section Answers

• Where does each air purifier brand stand in AI recommendation coverage, and what diagnostic should it prioritize?
• Which brands show stable coverage versus significant movement across the series?

The benchmark identifies where attention is warranted across the ten tracked brands; a company-level analysis is needed to explain why the observed patterns are occurring.

Evidence Behind the Benchmark

The aggregate metrics are built from prompt-level observations that capture the query, the AI surface, the recommendation outcome, the rank position, sentiment, and citations where exposed. Company-level analysis can go deeper into prompt, competitor, surface, and evidence patterns. Source presence is not automatically treated as proof of causation.

Report-Specific Interpretation Notes

• Small-count movement: Brands with low absolute valid recommendation counts can show percentage movement from small changes in the underlying count. Molekule's 1.0% coverage in October reflects 6 valid recommendations, and Honeywell's 11.4% reflects 70.
• Qualified denominator vs raw collection: All percentages use the 612 qualified October observations, not the 800 raw prompts collected. The July baseline used 615 qualified observations.
• Directional analysis: The benchmark's significance classification identifies which changes are large enough to track across months; it does not by itself establish the cause of those changes.

Next Step

The Public Benchmark Shows Where a Brand Is Winning or Losing. A Company-Level Audit Shows Why.

The aggregate percentages only reveal the surface. Beneath them sit the questions that matter: which high-intent prompts are won, which competitor takes the recommendation when a brand loses, what attributes AI associates with each option, and which external sources shape those answers.

Coway's three-month climb could stem from changes in a small number of influential evidence sources or from a broader shift across many prompts. Austin Air's and Dyson's declines could be concentrated in a handful of high-visibility contexts or spread across the full question set. The public benchmark cannot distinguish between these explanations. A company-specific AI visibility audit maps prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy.

Request an AI visibility audit