The vertical application scenarios of smart glasses are converging, and the goal of "all-day wearable" is difficult to be fully realized within the next 3 to 5 years.

At WAIC 2026, the smart glasses track sent a clear signal: the industry is shifting from "trying to do everything" to "figuring out what problems to solve for whom".

At the exhibition venue, new smart glasses products were unveiled in clusters, with tech giants like Alibaba, iFLYTEK, and Tencent competing on the same stage alongside specialized manufacturers such as Rokid, INMO, Liville, and LLvision, covering different scenarios including recording, translation, memory, and health, with long queues of experience seekers. Among them, Alibaba's Qwen launched the S1 series AI glasses, iFLYTEK introduced its first AI translation glasses, Liville partnered with Tencent Cloud to launch memory glasses, Moonix achieved a weight of 14.9 grams, and Rokid released the YodaOS operating system.

Data also confirms this popularity. According to Omdia statistics, global AI glasses shipments reached 8.7 million units in 2025, a year-on-year increase of 322%. Sigmaintell predicts that full-year shipments in 2026 will reach 16.1 million units, a year-on-year increase of about 72%; IDC forecasts that the Chinese market will exceed 4.915 million units in 2026, a year-on-year increase of 77.7%, officially entering the stage of large-scale growth.

However, behind the bustle, an awkward reality is gradually emerging: as some practitioners put it, 80% of the AI glasses on the market have almost identical internal components when disassembled, with extremely similar core configurations. As the supply chain rapidly lowers the achievable thresholds, competition is no longer about whether a product can be made, but why consumers would choose to buy this particular one.

JIA Hao, Vice Chairman and Secretary-General of the Zhongguancun Digital Intelligence Artificial Intelligence Industry Alliance, told *Caijing* New Media directly that the current smart glasses industry as a whole is in a stage where "segmented scenarios are taking the lead in proving viable, while mass market essential demands still need to be cultivated". A closed loop of rigid demand covering all groups of people has not yet been formed, but a clear "demand-payment-repurchase" business logic has been verified in vertical tracks.

Strategic Differentiation: Players Anchor on Distinct Scenarios

Compared to previous years that focused on technological breakthroughs and model iterations, at WAIC 2026, artificial intelligence is fully evolving into a "smart partner" featuring proactive services, human-machine collaboration, and cross-domain empowerment. Focusing on smart glasses, different players are beginning to answer the same question in different ways: for whom exactly do these glasses solve what problems?

In comparison, Liville is betting on "long-term memory". Its launched AI memory glasses are positioned as "memory extension". The X-AI memory glasses weigh only 26 grams, equipped with a self-developed memory strategy engine that can intelligently identify key scenarios, automatically record and generate structured minutes and to-do items.

In terms of privacy protection, Liville promises "processing only, no storage". The module can be removed when visual recording is not needed, reducing device weight while effectively avoiding shooting privacy risks in private scenarios such as offices and meetings, achieving a balance between functionality and privacy. At the same time, Liville also announced an ecological strategic cooperation with Tencent Cloud WorkBuddy, connecting the complete chain from offline perception to cloud execution, promoting AI Agent to move from online interaction to long-term services in real scenarios.

Qwen is betting on "Agent + Vital Sign Monitoring", upgrading AI glasses to "Agent Glasses", adding full-duplex voice, eye tracking, and the industry's first real-time user vital sign monitoring (heart rate, blood oxygen, HRV, etc.), which can generate personalized health reports and form Agent memories. After the upgrade, the glasses can enhance service and decision-making capabilities through the Agent, and call third-party Skills and Agents on demand.

BleeqUp, a brand under Unknown Tech, focuses on the "AI + Sports" track. At this WAIC, it fully unveiled its "AI Sports Health Smart Hub" strategy for the first time, and clarified its long-term ecological positioning for the first time. That is, taking BleeqUp smart glasses as the hardware entry, relying on the capabilities of a vertical sports AI Agent, linking the entire industry chain of sports equipment, health monitoring, and outdoor services to build an open AI sports health ecosystem covering global users.

What it aims at is trying to redefine the relationship between smart wearable devices and humans, allowing AI to actively understand scenarios, predict needs, and provide decision support at critical moments, achieving a leap from "quantification" to "optimization". From the perspective of WU Dezhou, founder of BleeqUp, end-side AI is not about running parameters in a laboratory, but making real-time decisions in outdoor scenarios such as cycling, running, mountain climbing, and skiing. The device not only records what you have done, but also determines what you should do, and even intervenes in advance when you are not aware of the risks.

LLvision showcased a concentrated AR+AI application system covering three major scenarios: personal communication, organizational collaboration, and enterprise productivity, including the Hey2AR translation glasses, "Whisperer" subtitle glasses, AR+AI conference translation system, GLXSSSE, LeionPro, and ARISE enterprise service platform, demonstrating the development trend of AR+AI evolving from single-point capabilities to systematic scenario capabilities.

iFLYTEK focuses on translation. Its iFLYTEK AI glasses weigh about 40 grams, equipped with multimodal noise reduction and lip movement recognition technologies, supporting 122 languages (including accents), covering modes such as side-listening simultaneous interpretation, face-to-face translation, online simultaneous interpretation, and call translation. It also has functions such as automatic meeting minute generation, real-scene information retrieval, and a portable smart assistant.

Moonix is betting on "unobtrusive recording + AI generation", committed to creating proactive AI glasses based on personal data. Its first product, the standard version of Moonix AI glasses, weighs only 14.9 grams, actively discarding functions such as translation, teleprompter, and navigation, focusing resources on "recording", with a combined battery life of up to 16 hours.

It is reported that Moonix is positioned as "your long-term AI companion that truly understands you", different from the "computing platform" route pursued by tech giants, so as to build competitive barriers. The product initially targets business people and knowledge workers, prioritizing the domestic market. GUO Yuchen, President of Xinmou Technology, the operating entity of Moonix AI glasses, said in an interview with the media that "recording" is a function that can generate time compound interest. Long-term accumulated data will form a personal exclusive knowledge base, while functions such as translation and teleprompter are used less frequently and cannot bring continuous value-added effects.

Rokid unveiled YodaOS, a native AI operating system specially built for smart glasses, aiming to enable AI glasses to realize a full-link intelligent closed loop of "perception-understanding-execution-interaction", rather than being an entry point for a passively responsive voice assistant. At the WAIC site, Rokid also demonstrated the full-scenario implementation results and global ecological layout covering consumer, cultural tourism, and industrial sectors, announcing that the smart glasses industry has officially entered a new era driven by native AI operating systems.

Uneven Commercialization of Scenarios: How to Complete the Business Closed Loop?

Clearly, the smart glasses industry is upgrading from a single "volume expansion by category" to "scenarios and entry points".

A research report by Huatai Securities believes that the current outbreak of tech consumption is essentially the result of deep coupling between underlying technologies (end-side AI, embodied intelligence, digital manufacturing, etc.) crossing the practical threshold and consumers' pursuit of efficiency improvement, emotional companionship, and personalized expression.

Among them, terminals such as AI glasses, smart wearables, service robots, and 3D printing are expected to become new entry points for consumption and new scenario carriers. If subsequent device penetration rates, service subscriptions, and ecological closed loops continue to be verified, the profit logic of future products is expected to extend from one-time hardware sales to continuous service monetization.

JIA Hao analyzed that on the consumer side, audio-visual entertainment and first-person AI shooting are currently the tracks with the highest commercial maturity. The user value of such scenarios is intuitive, and the product experience has differentiated advantages, which can provide experiences that are difficult for mobile phones to replace, such as large-screen immersive movie viewing and hands-free first-person shooting. According to market data, from January to May 2026, the average price of the domestic smart glasses online market was about 2041 yuan. Among them, products priced in the range of 1500-2000 yuan that focus on the above core functions achieved the highest sales efficiency with balanced cost-effectiveness.

On the industrial side, B-end scenarios such as industrial inspection, remote operation and maintenance, and medical assistance have verified clear value in reducing costs and increasing efficiency. Smart glasses can standardize operation processes and enable real-time collaboration of remote experts, forming stable project-based procurement demands in fields such as energy, manufacturing, and clinical teaching. However, restricted by factors such as high hardware procurement costs and long industry customization adaptation cycles, it is still dominated by benchmark pilots and has not achieved large-scale rigid demand penetration across the industry. There is still room for cost reduction to achieve inclusive implementation.

Regarding the widely discussed topic of "pseudo-demand" in the market, JIA Hao gave a relatively restrained judgment. He prefers to define it as "phased differences in demand matching". Some general public daily scenarios, such as general-purpose AR navigation and full-scenario portable AI assistants, still have insufficient experience maturity and functional irreplaceability in current products. They have not yet established users' habitual wearing mindset, and are still in the verification period of demand polishing and product iteration, requiring further deep integration of technology and scenarios to complete the business closed loop.

At the current stage, most consumer-grade smart glasses still exist in the form of mobile phone supporting terminals, relying on mobile phones to realize computing power supplementation and ecological linkage, which is the industrial status determined by the stage of technological development. However, products supporting independent eSIM internet access have emerged, and the industry is gradually exploring the form of independent terminals that are separated from mobile phones.

JIA Hao said that in the long run, with the continuous breakthrough of end-side computing power, optical technology, and battery technology, smart glasses with independent spatial computing capabilities have full potential to become the next ubiquitous computing entry after smartphones. This evolution process is gradual and cannot be achieved overnight.

Regarding the industrial evolution in the next 5-10 years, JIA Hao believes that in the next 3-5 years, the industry will still focus on vertical scenario implementation. B-end industrial scenarios will take the lead in achieving large-scale penetration, while the C-end will continue to focus on segmented groups, accumulating users in tracks such as audio-visual, AI creation, and accessibility assistance. From a 5-10 year perspective, with breakthroughs in core technologies, smart glasses are expected to penetrate into general consumer products, with lighter forms and more comprehensive functions, gradually entering the mass consumer market.

"All-day Wearing" May Be Difficult to Implement in 3-5 Years, Enterprises Explore Their Own Paths

However, regarding the ultimate form of "all-day wearing", JIA Hao believes that the industry still has a long way to go in tackling key problems, and it is difficult to achieve popular all-day wearing implementation in the short term of 3-5 years.

The core bottlenecks come from two aspects: first, the underlying hardware constraints. The technical bottlenecks of battery energy density and low-power computing chips have not been broken through, and the balance between power consumption and functions in complex outdoor environments is still a common problem across the industry. Second, user value constraints. Most of the current functions of smart glasses can still be replaced by mature terminals such as mobile phones and tablets, and irreplaceable value that users must wear all day long has not yet been formed. The establishment of the public's wearing mindset still takes time.

It is reported that the current technological progress in the smart glasses industry is mostly engineering optimization and gradual iteration, and there has not been a subversive breakthrough at the underlying principle level, which is also the consensus of the whole industry.

JIA Hao said frankly that in the field of optical display, the progress of mainstream solutions focuses on yield improvement, cost reduction, and volume optimization. The industrialization implementation speed of technologies such as optical waveguides and Micro LED is accelerating, but there has been no essential change in the underlying optical architecture. The core is still to find a balance among display effect, weight, and cost.

In the field of end-side AI computing power, the core breakthrough is the lightweight end-side deployment of large models, realizing the local operation of functions such as AI shooting, real-time translation, and voice interaction. Essentially, it is the result of AI engineering implementation, not a generational upgrade of the computing power architecture. At present, end-cloud collaboration is still the main computing power mode.

In the field of lightweight design, the current weight optimization of products is more about "subtraction" through function trimming and structural integration, prioritizing the wearing experience of core functions under the constraints of limited battery and computing power, rather than revolutionary weight reduction of underlying hardware.

In fact, as an integrated product spanning multiple fields such as hardware, optics, AI, content, and industry applications, smart glasses continue to attract tech giants, startups, and industry chain enterprises to enter the market. Market entities with different backgrounds cut in based on their own resource endowments, forming a parallel pattern of multiple camps, and opening up their own segmented tracks in this "hundred-glasses competition". Capital is also spreading to both ends of the industry chain along with the track boom, intensively betting on terminal brands on one hand and increasing investment in the upstream supply chain on the other.

JIA Hao believes that the current differentiated exploration of domestic smart glasses enterprises is an inevitable choice for the industry to develop to this stage. All enterprises find entry points for segmented tracks based on their own advantages. There is no absolute distinction between good and bad, and they jointly broaden the development boundary of the industry.

In comparison, enterprises such as XREAL and Thunderbird Innovation focus on the consumer-grade audio-visual track, deeply cultivating display experience and movie viewing scenarios, accurately capturing the clearest user demands in the current C-end market, forming clear product recognition in the consumer market. Enterprises such as INMO focus on fashionable design and innovation in segmented functions such as translation, exploring the breakthrough of the daily wear attribute of smart glasses and expanding the user circle of products.

A group of enterprises including NOLO deeply cultivate B-end vertical industrial scenarios, focus on solving actual industry pain points, drive commercial implementation with scenario value, and embark on a differentiated development route that combines To B and To C. These explorations have accumulated multi-dimensional practical experience for the industry, which is a necessary process for the industry to move towards maturity.

As for "who will become the Apple equivalent of China's smart glasses", JIA Hao admits frankly that the industry is still in the early stage of development, and the technical routes, user demands, and market structure are far from being finalized. It is too early to predict the final industry leader. The enterprises that can lead the industry in the future will definitely be market entities that have the capabilities of core technological breakthrough, ecological construction, and user value creation at the same time.

Before the arrival of the "iPhone Moment", a more pragmatic question is: who will first prove the profitable model of vertical scenarios, and who will first make users form the mindset of having to wear glasses every day.

This article is from "*Caijing* New Media", author: LIU Fen, editor: JIANG Shizhou, published with authorization by 36Kr.