Asklear ResearchUsing JD and Tmall's platform-defined “smart glasses” categories as a retail proxy, AI glasses are no longer only a concept story. In the second quarter of 2026, both category GMV and unit-sales research indicators increased on both platforms. The more precise conclusion is that an observable retail growth signal has emerged, while the current evidence does not establish mass adoption: JD showed the clearer month-by-month expansion, the platforms followed different price paths, and second-quarter GMV was concentrated among a small group of leading brands.
That distinction matters because global market trackers often start with shipments, new entrants, form factors, and possible use cases. Operators and consumers need a narrower answer first: are people buying these products at retail, does growth persist beyond a launch or promotion, and is demand spreading across brands and models?
JD and Tmall data can address part of this question and reveal the current brand structure. The platform category cannot establish that every included listing has AI functionality, whether buyers keep wearing the glasses, or whether orders remain after returns. This analysis therefore examines the signals produced when the smart-glasses category is used as an AI-glasses retail proxy: incremental transactions, continuity, and the breadth of the current GMV distribution.
“Really selling” is one layer of the commercialization chain
The phrase “really selling” can refer to shipments, retail orders, post-return paid transactions, or sustained usage. These layers are connected, but one cannot stand in for another. Global market trackers usually describe supply expansion, vendor share, and product form. Those measures establish industry momentum; they do not directly show how retail orders are changing on specific Chinese e-commerce platforms.
Asklear data sits at the retail-transaction layer. GMV, units, and ASP can establish whether a platform category generated more transactions and can be decomposed by month and brand. Compared with a market forecast, this advances the question from “How large might the category become?” to “Is observable retail activity already increasing?”
The scope is deliberately narrower than adoption. JD and Tmall use their raw level-three “smart glasses” categories, which may contain different product forms, and the current package does not classify AI capabilities listing by listing. Transacted units are also not post-return retention, activation, wearing time, or repeat usage. The article uses “AI glasses” because that is the reader's question, while the measured object remains the platform-defined smart-glasses category.
That proxy is still useful. If the retail category showed no transaction growth, claims of broad adoption would be difficult to sustain. If transactions are increasing, the next task is to assess their quality. Retail growth is therefore a necessary signal in the commercialization chain, not proof that the chain is complete.
Both platforms showed unit and GMV growth
The first test is not how large the market might eventually become. It is whether GMV and unit sales increased during the same period. Under Asklear's research definition, second-quarter smart-glasses GMV increased 29.7% from the first quarter on JD while units rose 47.2%. On Tmall, GMV rose 27.5% and units rose 23.5%. The shared direction shows that the retail growth signal was not isolated to one platform.
| Platform | Quarterly GMV change | Quarterly unit change | Quarterly ASP change |
|---|---|---|---|
| JD | +29.7% | +47.2% | -11.9% |
| Tmall | +27.5% | +23.5% | +3.2% |
Data source：https://asklearai.comView methodology
- Query scope
- JD.com · Tmall · 2026-01-01—2026-03-31 / 2026-04-01—2026-06-30 · Level-3 category「智能眼镜」
- Metric and grouping
- GMV research metric / Units research metric / Average selling price · Aggregate
- Calculation
- Change rate = (current period ÷ previous period − 1) × 100%.
- MCP query
query_metrics({ "dataset": "jd", "time": { "start": "2026-01-01", "end": "2026-03-31" }, "filters": [ { "field": "category_l3", "op": "eq", "value": "智能眼镜" } ], "metrics": [ "gmv", "units", "asp" ], "group_by": [], "order_by": [], "limit": 100 })query_metrics({ "dataset": "jd", "time": { "start": "2026-04-01", "end": "2026-06-30" }, "filters": [ { "field": "category_l3", "op": "eq", "value": "智能眼镜" } ], "metrics": [ "gmv", "units", "asp" ], "group_by": [], "order_by": [], "limit": 100 })query_metrics({ "dataset": "tmall", "time": { "start": "2026-01-01", "end": "2026-03-31" }, "filters": [ { "field": "category_l3", "op": "eq", "value": "智能眼镜" } ], "metrics": [ "gmv", "units", "asp" ], "group_by": [], "order_by": [], "limit": 100 })
两平台均使用原始三级类目“智能眼镜”，时间范围与指标定义一致；但平台商品、品牌、商家和促销覆盖不同，因此只比较趋势、结构和方向，不合并成交额或销量。平均成交价是平台内品类汇总值，不应被解释为同一商品的跨平台价差。Asklear research metrics; not platform, merchant-backend, refund-adjusted, or settlement truth.query_metrics({ "dataset": "tmall", "time": { "start": "2026-04-01", "end": "2026-06-30" }, "filters": [ { "field": "category_l3", "op": "eq", "value": "智能眼镜" } ], "metrics": [ "gmv", "units", "asp" ], "group_by": [], "order_by": [], "limit": 100 })- Data updated
Concurrent growth in units and GMV rules out the strongest version of the “attention without transactions” argument. It also makes a single-platform anomaly less plausible. The platform totals are not directly comparable or additive because the product, merchant, and promotion coverage differs; the useful cross-platform result is that both moved in the same direction.
Quarter-over-quarter growth is not automatically a durable trend. Launch schedules, promotional events, inventory availability, and category mapping can all affect the second quarter. The result establishes expansion within the observed half-year; it does not imply that future quarters will repeat the same growth rate. Price-volume composition and monthly continuity must carry the next stages of the argument.
Similar GMV growth concealed different price-volume paths
The relationship among GMV, units, and average selling price is more informative than the headline increase. JD's unit growth outpaced GMV growth while ASP fell 11.9%. Tmall's GMV growth slightly exceeded unit growth while ASP increased 3.2%. Growth in the smart-glasses proxy categories was therefore not one uniform discount-led expansion: JD leaned more heavily toward volume, whereas Tmall expanded units while maintaining a higher transaction-price mix.
ASP describes each platform's product mix, not a price gap for identical items. Promotions, launch timing, product functions, and the mix of price bands could all influence it. The evidence supports different within-platform mix changes; it does not establish that JD shoppers prefer cheaper devices or Tmall shoppers prefer premium ones.
For operators, that distinction changes the analysis. Category GMV should be read with units and ASP because higher GMV can come from more products being purchased or from higher-priced products taking a larger share. Those paths call for different product, inventory, and acquisition decisions.
The competing explanations are testable with deeper evidence. If JD's ASP decline came mainly from broad discounting, the same core products should show lower transaction prices and volume may fall after the campaign. If entry-level products expanded, lower price bands and their leading listings should retain a larger share. If Tmall's ASP increase came from premium launches, growth should be visible in those price bands and new listings; if lower-priced products were temporarily unavailable, the mix could reverse when supply returns.
ASP is therefore a research entry point rather than a consumer-preference conclusion. It tells the analyst that the mix changed, then points toward the price-band and product queries needed to explain why. The current package does not include those before-and-after structures, so the next investigation should add them instead of repeating the completed quarterly total queries.
JD's volume increase persisted month by month; Tmall dipped and recovered
Quarterly growth answers whether incremental transactions existed. The monthly path begins to answer whether they were durable. If a single promotion or launch produced the whole increase, the quarter would provide weaker evidence of sustained commercialization.
JD smart-glasses units increased in every month from April through June, finishing 49.9% above April. Three consecutive moves in the same direction make the second-quarter result more informative than one aggregate total: the transaction indicator did not immediately fall after the first growth month.
The chart should be read for the continuity from April through June, not only for the June endpoint.
JD smart-glasses unit sales rose in every month of Q2
Shows monthly transacted units from April through June to test whether quarterly growth came from one isolated month.
Data source：https://asklearai.comView methodology
- Query scope
- JD.com · 2026-01-01—2026-06-30 · Level-3 category「智能眼镜」
- Metric and grouping
- Units research metric · Monthly
- MCP query
Units are Asklear research indicators for JD's raw level-three smart-glasses category.Asklear research metrics; not platform, merchant-backend, refund-adjusted, or settlement truth.query_metrics({ "dataset": "jd", "time": { "start": "2026-01-01", "end": "2026-06-30" }, "filters": [ { "field": "category_l3", "op": "eq", "value": "智能眼镜" } ], "metrics": [ "gmv", "units", "asp" ], "group_by": [ "month" ], "order_by": [ { "field": "month", "dir": "asc" } ], "limit": 20 })- Data updated
- Chart values
- 2026-04 34,383 units；2026-05 42,722 units；2026-06 51,546 units
The rising path strengthens the continuity finding, but it cannot isolate demand from supply or marketing. More launches, better availability, and deeper promotions could all produce consecutive increases. The evidence establishes continuity in the result, not a single cause.
Promotions may also affect this period, so the rising line should not be labeled organic demand. The more discriminating evidence will come afterward: whether the following month preserves a higher baseline and whether the same listings continue to transact without an equally deep promotion. Continuity during the observed quarter is encouraging; continuity afterward would be more diagnostic.
Tmall followed a different shape. Units were 13,178 in April, fell to 10,834 in May, and recovered to 13,477 in June. June finished slightly above April, but the intervening decline makes “a larger quarter with monthly volatility” more accurate than “continuous acceleration.”
Together, the platforms support a more useful conclusion than a broad claim of an industry-wide breakout. The smart-glasses proxy categories generated incremental retail transactions, but continuity is currently clearer on JD and still requires more months of confirmation on Tmall. For operators, the next informative measure is not whether another campaign can create a peak, but whether the post-campaign sales floor keeps rising.
Second-quarter GMV was concentrated among leading brands
After continuity, the next question is how broadly current transactions are distributed. A category may have a few strong brands capture most GMV or show visible shares across a wider field. The second-quarter cross-section shows the former structure.
In the second quarter, the top five brands represented 69.4% of smart-glasses GMV on JD and 82.2% on Tmall. Current GMV was therefore concentrated among leading brands on both platforms. The dataset does not contain first-quarter brand shares, so it cannot show whether concentration increased or decreased, or which brands contributed the quarterly increment.
The gap within JD's leading group is also meaningful. ROKID led clearly in the second quarter, Huawei and Qianwen formed the next tier, and INMO and Quark followed. The chart carries the exact magnitude comparison; the analytical point is that GMV was uneven even within the leading group.
The chart compares the internal structure of JD's top five brands; it does not compare JD's absolute scale with Tmall.
ROKID held a clear GMV lead within JD's top smart-glasses brands
Compares second-quarter GMV among JD's top five brands to show the magnitude gaps within the leading group.
Data source：https://asklearai.comView methodology
- Query scope
- JD.com · 2026-04-01—2026-06-30 · Level-3 category「智能眼镜」
- Metric and grouping
- GMV research metric · Brand
- MCP query
JD only. Huawei combines two confirmed raw brand aliases.合并平台原始品牌值“华为（HUAWEI）”与“华为/HUAWEI”。Asklear research metrics; not platform, merchant-backend, refund-adjusted, or settlement truth.query_metrics({ "dataset": "jd", "time": { "start": "2026-04-01", "end": "2026-06-30" }, "filters": [ { "field": "category_l3", "op": "eq", "value": "智能眼镜" } ], "metrics": [ "gmv", "units", "asp" ], "group_by": [ "brand" ], "order_by": [ { "field": "gmv", "dir": "desc" } ], "limit": 20 })- Data updated
- Chart values
- ROKID 69,255,086.25 CNY；华为 31,652,413.18 CNY；千问 26,518,432.16 CNY；INMO 19,774,725.36 CNY；夸克 14,333,723.25 CNY
This concentration supports the conclusion that the category had identifiable leaders in the quarter, but one quarter cannot prove that the competitive structure is fixed. Product launches can reorder brands, and alias or category mapping can affect rankings. More importantly, the data identifies who captured GMV; it does not establish why consumers chose them.
For a new entrant, category growth alone is not a sufficient entry case. The next questions are which price bands and products carry the leaders, whether non-leading models preserve volume after campaigns, and whether incremental transactions expand the category or redistribute existing demand. For investors, concentration is also a warning against extrapolating the success of leading brands to every participant in the AI-glasses supply chain.
Different readers should take different next steps
The same evidence should not produce one generic bullish or bearish verdict. Its value is to narrow the next decision for each audience.
Brand and e-commerce teams need to separate growth sources. JD's units rose faster than GMV, so the next analysis should examine lower price bands, promotion depth, and the post-event sales floor. Tmall's ASP increased slightly, so operators need to identify the price bands and listings supporting the higher mix. Tracking category GMV alone would combine distinct operating conditions.
Investors can support the statement that observable retail transactions increased, but not that consumer adoption is complete. A fuller commercialization view must connect platform orders with returns, activation, wearing frequency, retention, and channel inventory. Commerce data answers a consequential step in the demand chain; it is not the endpoint.
Publishers should be especially careful about metric substitution. Global shipments, domestic retail transactions, platform settlement, recognized revenue, and actual use answer different questions. Citing one growth rate can easily turn supply expansion into consumer adoption or a leading brand's performance into an industry-wide uplift. Naming the layer being measured is part of the article's credibility.
What evidence would change the conclusion?
The current conclusion—observable transactions have emerged, while mass adoption remains unproven—must be falsifiable. Later evidence could move it in several directions.
The conclusion would weaken if post-promotion units returned to the previous baseline, growth remained confined to a few event months and listings, and ASP changes reversed as incentives ended. JD's consecutive second-quarter increases reduce the likelihood of a one-month spike, but they do not yet cover the period after the promotion.
The conclusion would strengthen if later months preserved a higher unit floor, stable transactions appeared across more price bands and products, and a broader set of brands captured meaningful shares. Those questions require before-and-after structures; a single-quarter brand ranking cannot answer them.
The analysis would move from transactions toward adoption only if post-return orders, device activation, usage frequency, and retention increased alongside commerce metrics. That evidence must come from outside the current e-commerce dataset. Adding more GMV charts would not close the gap.
Transactions are established; mass adoption is not
Taken together, the platform-defined smart-glasses categories provide evidence that AI glasses can no longer be described as attention without observable retail transactions in the first half of 2026. GMV and unit-sales indicators increased on both platforms, JD expanded in every month from April through June, and the top five brands captured most second-quarter GMV on both platforms.
The evidence does not yet establish mass adoption. Tmall units did not increase every month; ASP moved in opposite directions on the two platforms, suggesting different product mixes; and second-quarter GMV was concentrated among a few brands.
The most accurate answer to “Are AI glasses really selling?” is therefore: the smart-glasses proxy categories show observable retail growth; the paths differ by platform, second-quarter GMV is concentrated among leading brands, and current evidence does not establish mass adoption.
Methodology
This analysis uses Asklear monthly JD and Tmall e-commerce research data from January through June 2026 for each platform's raw level-three “smart glasses” category.
The platform category serves as a retail proxy for AI glasses in this article and cannot identify the AI functionality of every included listing. GMV is a post-coupon sales-value research indicator, units are a transacted-item research indicator, and ASP is the platform-level category average. These measures are not platform settlement, post-return payment, or recognized revenue. Confirmed brand aliases were consolidated. Because platform product, seller, and promotion coverage differs, comparisons are directional and structural; absolute totals are not combined.