Why we're looking at this brand again
A previous article on this hot topic reported that "Qwen," "Quark," and their Chinese-language equivalents scored zero mentions in our 2026-09-02 run. This isn't a repeat of that finding. Two things changed since then. First, we re-ran the same question after fixing a brand-normalization pass, and this article cites that later run (2026-09-03 10:25:24), not the one the earlier article used. Second, we verified a fact we hadn't written up before: "Qwen Glasses" is not a rebrand of "Quark AI Glasses" — it's a separate product Alibaba launched afterward. Put together, that's more interesting than the zero-mention number on its own.
Snapshot of the measurement
In the 2026-09-03 10:25:24 run, our English-language panel returned 25 candidate brands with full panel completeness. "Qwen," "Quark," and their transliterations scored zero — not low, absent from the list entirely. The top three were XREAL (66.44), Rokid (62.51), and Ray-Ban (60.31).
One data point worth calling out on its own: the panel includes Alibaba's own model, qwen/qwen3.8-max. Asked (in the Chinese-language run, same measurement family) which AI glasses brands were most popular, it named ten: Ray-Ban, Meta, Xreal, Rokid, RayNeo (Chinese: 雷鸟创新), INMO, Lyeekaka (李未可), Huawei, Xiaomi, and Even Realities. Neither Qwen nor Quark is on that list. We're reporting the raw response, not speculating about why.
Observation one: a rename story doesn't quite hold up
The following background on Alibaba's product timeline is a third-party citation we have not independently verified; treat it accordingly. Per South China Morning Post reporting dated 2026-02-28, Alibaba launched "Quark AI Glasses" in November 2025, then separately unveiled "Qwen Glasses" at MWC Barcelona on 2026-03-02/03 — described in that report as Alibaba's "second" entry into smart eyewear, not a renamed continuation of the first.
That timeline partly explains the zero-mention result, but not fully. Several head models in this panel report training cutoffs clustered between January and March 2026 — plausibly too early to have absorbed "Qwen Glasses" as a name. But "Quark" has been in use since November 2025, well before those cutoffs, and it still scored zero. "The name is too new" doesn't account for that half of the gap.
Observation two: the gap between shipment share and mindshare hasn't closed
The following market-share figures are cited from a third-party research firm, CINNO Research; we have not independently verified them, and accuracy is the publisher's responsibility. CINNO's Q1 2026 estimate of China's AI/AR glasses market puts RayNeo at 24%, Rokid at 20%, XREAL at 14%, Qwen at 11%, and Xiaomi at 7% — placing Qwen ahead of Xiaomi by that measure.
Set the two datasets side by side: Rokid and XREAL show up in the top five of both CINNO's shipment-share estimate and our mention-rate measurement. Qwen shows up in the top five of the shipment data and at zero in ours. The two measurements aren't the same kind of thing — one counts units moved through retail and e-commerce channels, the other counts what a model's training corpus produced when asked a question — so they can't confirm or substitute for each other. We're not calling either one more accurate. We're reporting that the gap exists.
Limits
Panel completeness for this run was 100%, with 7 head models plus several long-tail models weighted per our methodology page. Normalization status was ok. One thing worth flagging: our evaluation panel is itself made of large language models, and those models' training data may include the results of other AI rankings and reviews — a form of self-reinforcing contamination where a brand's presence partly reflects having been mentioned elsewhere before, not fully independent judgment. That's a known limitation of the current methodology, not something specific to this run.
Two more limits specific to this article: the SCMP product-timeline reporting and the CINNO shipment-share figures are both third-party citations we have not independently verified. And a zero-mention score is not a judgment of the Qwen Glasses product or of Alibaba's marketing — it only describes what the training corpora of the queried models currently contain.
Full methodology, every raw model response, and the normalization log are at our methodology page and full rankings. Two related pieces on this topic: AI Glasses: Every model puts Ray-Ban first, but it ranks fourth after aggregation and One brand near the top of China's e-commerce sales charts scored zero mentions across 12 models.