Smart Glasses in 2025: Path Divergence Behind High Hype and Low Sales
In 2025, the smart glasses market has become incredibly crowded and bustling. AI audio glasses, AR glasses, and various scenario - focused glasses have emerged in an endless stream, becoming a "testing ground" for enterprises to explore the next - generation interaction.
A seemingly simple pair of glasses is drawing global tech giants and local innovative enterprises into a complex war regarding interaction, entry points, and ecosystems. In 2025, the smart glasses market has become incredibly crowded and bustling. AI audio glasses, AR glasses, and various scenario - focused glasses have emerged in an endless stream, becoming a "testing ground" for enterprises to explore the next - generation interaction.
Overseas, Meta continues to lead with an extremely high market share, while Google and Apple are making aggressive moves. In China, internet and hardware giants such as Alibaba, Xiaomi, and Baidu are leveraging their advantages in ecosystems, funds, and supply chains to enter the market comprehensively. Their new products have quickly made it onto the best - selling lists. The full - scale entry of giants is rapidly squeezing the time window for startups to define products, educate the market, and build barriers.
However, a consensus and a mystery coexist. The consensus is that almost all industry players view the next two to three years as a crucial period of explosive growth, and it is essential to secure a position in the market. The mystery is that so far, there has not been a signature blockbuster product in China with an annual shipment volume exceeding 500,000 units. The market is still waiting for a product that truly meets the public's needs. Therefore, "choosing the right direction" is more important than blindly chasing after trends. Many entrepreneurs point out that the current core strategy is to focus on a vertical, concrete scenario with real pain points, dig deep into it, and strive to make the product a synonym for the niche category, thus establishing a strong brand association in consumers' minds.
"No matter what kind of smart glasses people wear, the first question enterprises need to answer is: What problems can it solve?" Wu Fei, the founder and CEO of LLVision, told Caijing New Media. He cited the classic example of the first - generation iPod, which aimed to "put 1000 songs in your pocket," and pointed out that great products often start with an extremely simple and clear answer. History doesn't repeat itself exactly, but patterns often reappear.
Facing the strategic crossroads, should companies build a general - purpose platform or a vertical tool? Should they focus on display interaction or audio? Should they pursue ultimate performance or prioritize battery life? Different companies have given different answers. Behind these choices lies the judgment of technological trends and a bet on market demand, testing the vision and decision - making ability of the leaders.
Fast: Giants are encroaching on all fronts, and their products are already on the lists
Recently, the competitive landscape in the smart glasses field is being rapidly reshaped globally.
Overseas giants have entered a new stage with a clear strategy. Meta has confirmed that it will shift its resources towards AI glasses and wearable devices. Google has returned to the market in a high - profile manner, announcing four XR product lines at once, aiming to compete directly with Meta in 2026. It is reported that Apple regards AR glasses as its "highest strategic priority." Tim Cook is determined to create a benchmark product that surpasses Meta's. The first smart glasses are expected to be unveiled by the end of 2026.
The domestic battlefront shows characteristics of multiple paths and multiple ecosystems, and the main players can be divided into three categories.
The first category is internet giants such as Alibaba, Baidu, ByteDance, and Tencent. The essence of their competition is a battle for entry points. Relying on their large - model capabilities, they quickly launch hardware products, aiming to build a full - stack ability of "model - application - hardware" and compete for the right to speak in the data of the physical world and the entry points in the AI era.
The second category includes mobile phone manufacturers represented by Xiaomi, Huawei, OPPO, and Xingji Meizu, new - energy vehicle companies represented by Li Auto, and telecom operators represented by China Telecom and China Mobile. Their core logic is to extend their ecosystems.
The third category is innovative smart glasses enterprises in the virtual reality field, such as Thunderbird, Rokid, XREAL, INMO, and LLVision. They rely on their technological accumulation and rapid iteration ability in the VR/AR/MR fields to accelerate the layout of the industrial chain and innovation on the application side, so as to promote large - scale mass production. There are also leading hardware enterprises such as Dreame and ZOTO, which are venturing into new forms of smart hardware.
Despite the grand momentum, the industry is still in the early stage of "high popularity, low sales." IDC data shows that including all kinds of products such as AR glasses with displays, audio glasses without displays, and shooting glasses, the global shipment volume of smart glasses in the third quarter of 2025 increased by 74.1% year - on - year, but Meta alone accounted for 75.7% of the market share. Although four Chinese manufacturers, Xiaomi, Thunderbird, Xreal, and Viture, made it into the global top five, their market shares were all less than 5%. Although the domestic market showed rapid growth during the "Double 11" period, according to industry insiders' observations, the overall annual shipment volume is about one million units, which is still less than Meta's expected single - company sales volume. Moreover, there is still a lack of a signature blockbuster product with a shipment volume exceeding 500,000 units in the market.
As of December 28, on JD.com's hot - selling list of smart glasses, Alibaba's Quark AI Glasses S1 has ranked first for 17 consecutive days, and the G1 has ranked eighth. ROKID Glasses, Xiaomi AI Glasses, Li Auto's AI Smart Glasses Livis, and INMO's INMOGO3 AR Smart Glasses ranked second to fifth respectively. Thunderbird's V3, Jiehuan AI Glasses, and Meizu's StarV Air2 Smart Glasses also made it into the top 15 of the list.
On Tmall's hot - selling list of smart glasses, ROKID Glasses ranked first in sales and has dominated the top three for eight consecutive weeks. Products from INMO, Thunderbird, and LLVision ranked second, fourth, and seventh respectively. Among the top ten products, those marked with Xiaomi and Xiaodu accounted for four positions.
Slow: Blockbuster products have not emerged yet, and direction is more important than speed
In contrast to the rapid position - securing of tech giants, the popularization of products has to go through a "slow - grinding" process. A widely envisioned ultimate form of XR is a pair of glasses as light as ordinary ones, capable of seamlessly integrating into daily life and achieving the seamless overlay of virtual and real information.
Compared with MR and VR, AR glasses currently have an edge in this form competition. According to the latest statistics from CINNO Research, in the third quarter of 2025, the sales volume of consumer - grade AR glasses with screens in China increased by 355% year - on - year. The key to this breakthrough is that core components such as batteries, chips, and optical modules have been highly integrated and successfully "compressed" into the glasses form, allowing users to obtain real - time information feedback while seeing the real world.
However, stuffing technology into glasses is just the starting point. Many interviewees pointed out that a pair of smart glasses must first be a good pair of glasses. The essence of the product is what functions are added to the glasses to better balance user needs and technological implementation.
In 2025, many display smart glasses have achieved a lightweight level of about 50 grams, approaching that of an ordinary pair of sunglasses. Shargeek's loomos S1 claims to be "possibly the lightest AI glasses in the world," with a total weight of 29 grams and a perceived weight of only 15 grams. Audio glasses products from brands such as Jiehuan can weigh about 26 grams (without lenses), close to the weight of an ordinary spectacle frame.
To achieve this weight reduction, a comprehensive game of technology, balance, and resources is required. It involves the ultimate breakthrough in optics, materials, and structure, precise control of performance and battery life, and deep support from technological precipitation, top - tier suppliers, financial strength, and capital resources.
Take the binocular AR glasses X3 Pro as an example. Thunderbird uses the world's smallest mass - producible full - color Micro - LED firefly light engine, self - develops an optical engine that is not available on the market, and jointly develops optical waveguides with Applied Materials in the United States. Eventually, the weight is reduced to 76 grams. In terms of display experience, it not only benefits from Thunderbird's optical and supply - chain precipitation but also from TCL's empowerment, including an understanding of display and color.
The deeper challenge lies in the high fragmentation of the industry's supply chain. The lack of unified standards means that core components such as optical waveguides and Micro - OLED screens cannot be used interchangeably, and manufacturers need to customize them for each brand. This drives up costs, lengthens the production cycle, and fundamentally limits the improvement of overall production capacity.
In the view of He Xiaomin, an analyst at WK Securities, the current competition logic in the consumer - grade AR glasses market will follow a "gradual iteration" path, evolving from short - term hardware competition to long - term ecosystem construction.
Facing these systematic challenges, innovative enterprises have shown significant differences in their path selection based on their judgments of the future.
Firstly, some enterprises aim to build a general - purpose "AR platform," such as Thunderbird and INMO Technology. According to the "Report on the Global AR Smart Glasses Brand Shipment Share in Q3 2025" released by Counterpoint Research, the Chinese brand Thunderbird Innovation's RayNeo has ranked first in the global AR smart glasses market for two consecutive quarters with a market share of 24%.
Yang Kexiao, the hardware director of Thunderbird, told Caijing New Media that although Thunderbird values the To C market, its long - term goal is to build a standardized open platform similar to Apple's. By building an open SDK, practitioners in different industries can directly use the device to meet specific needs, just like using an iPad to order food. Currently, Thunderbird's product logic is to start from vertical scenarios with strong experiences such as movie - watching and gaming, and then naturally expand to the general public after iterative improvement.
Secondly, some enterprises aim to create a "champion product" in a vertical niche, represented by LLVision and Zhijing Weizhi.
Contrary to the platform - oriented thinking, LLVision focuses all its industrial - grade AR technology accumulation on the AR translation glasses. The goal is to make it one of the most commonly used translation tools globally in the next three years. Wu Fei, the founder and CEO of LLVision, believes that the next 3 - 5 years will be a crucial window period for "vertical champions" in the XR industry. The company's strategic path is to first perfect the "face - to - face translation" scenario, build barriers with definite value, and then expand to a broader range of smart glasses capabilities.
Thirdly, some enterprises view AR glasses as a natural extension of consumer electronics. For example, TOZO and Dreame, which enter the market from the mature hardware field, focus on sports and health monitoring when expanding their smart wearable product lines.
"The XR industry is in an upward period of smart hardware popularization and content explosion. With the increasing maturity of 5G and the approaching of 6G, and the support of AI, consumer - grade XR devices will explode in various scenarios and niches. Now it depends on who can seize this opportunity," Wu Dajian, the product director of TOZO, told Caijing New Media.
TOZO is a smart wearable brand founded in 2015, with a global brand development strategy and its R & D center in Shenzhen. The brand has witnessed the explosion and popularization of TWS earphones and has long occupied the top positions on Amazon's TWS earphone list. It aims to create truly immersive and integrated smart wearables.
Based on his hardware manufacturing experience, Wu Dajian believes that the system and process of making glasses are in line with those of making mobile phones. In comparison, these enterprises, relying on their mature hardware productization, rapid iteration, and supply - chain management, select and delve into a niche market based on clear consumer needs, and enter the market with their hardware efficiency advantages, aiming to replicate their success in the consumer electronics field.
Choose: Anchor on "real needs" and bet on the future respectively
Behind the three models are different bets on when the future will arrive and a clear understanding of their own core capabilities. The industry consensus is that whether it is the dream of building a platform or becoming a champion, one must start from a specific scenario with deep - seated and painful problems to find the key to the mass market.
Amid the frenzy of the evolution of smart glasses functions, a discussion about real needs and false needs is also underway.
Skeptics believe that many flashy functions are false needs. Li Nan, a former senior executive of Meizu, publicly stated that "all display applications are false needs" and claimed that surveys showed that functions such as navigation and payment, which are highly popular, fall into this category.
Fengchao Technology is an enterprise in Xiaomi's ecosystem. Its founder, Xia Yongfeng, once frankly told Caijing New Media that he is not optimistic about categories with cameras or displays in the Chinese market. In his view, innovative hardware should replace old things rather than create new habits. Based on the large number of myopic people in China, he advocates that smart glasses should first follow the path of replacing optical glasses. Therefore, his company's products focus on high - cost - performance AI audio glasses.
However, manufacturers deeply involved in specific scenarios have given different answers through practice. Wu Fei of LLVision believes that as long as a function is truly usable, it may be a "major need." They initially developed AR subtitle glasses for the hearing - impaired, and later found that the need for real - time understanding extends beyond the hearing - impaired to a broader face - to - face conversation scenario.
He said that the company focuses on "face - to - face translation" because this kind of need is extremely demanding in terms of immediacy and specialization: it requires low latency, allows people to look at the interlocutor all the time, and requires continuous subtitles. To make "seeing sound" a truly usable and good - to - use tool, AR glasses are the closest to meeting these conditions in the current form.
The actual market choices show a practical differentiation. IDC predicts that in 2026, audio and shooting products will account for the absolute majority in the shipment volume of smart glasses in China, exceeding 3.4 million units, while AR/VR devices will be about 1.07 million units. This reflects the market's preference for lightweight and easy - to - use products under the current technological conditions.
The strategies of mainstream manufacturers are not mutually exclusive but involve a two - line layout. For example, Thunderbird and Rokid are leveraging their advantages to develop AR glasses with displays to explore the next - generation terminal form on one hand, and on the other hand, they are also deploying glasses without displays but focusing on AI shooting to accurately meet the immediate need for instant recording in scenarios such as sports and parenting.
Yang Kexiao analyzed that the core reason why Thunderbird was able to launch AI glasses more than half a year ahead of the industry is that it regards AI glasses as a technological subset of AR glasses with displays. This means that the company can modularize and lightweight the core technologies that have been proven successful in AR products in advance and quickly derive independent AI glasses products, thus achieving an efficiency advantage in technology reuse and product iteration.
Thunderbird's path is to start from niche but strongly - demanded scenarios such as movie - watching and gaming, which facilitates rapid iteration in the early stage of technology and accurate response to user feedback to verify the core experience. He believes that although the current technology is limited in terms of screen size, after continuous refinement in vertical scenarios, the niche needs will naturally expand into strong mass needs in the next few years when the technology matures.
Breakthrough: Penetrate vertical scenarios and break the deadlock with ecological synergy
To make smart glasses transition from geek toys to daily use by the general public, the key lies in penetrating a specific scenario and rallying ecological forces to jointly break through the impossible triangle of performance, weight, and battery life. Many interviewees said that the development of the industry requires the collaboration of industrial clusters, integrating the forces of the government, supply chains, enterprises, and universities. Currently, the industry logic is becoming clear: define product standards through extreme scenarios and force the evolution of the supply chain.
Take LLVision as an example. Facing the seemingly impossible "want - it - all" problem of "fast, accurate, long - lasting, and lightweight" in the AR translation field, the company did not make patches within the existing framework but chose to reconstruct the technology. It achieved form weight reduction through a thinner diffractive optical waveguide, reduced power consumption step - by - step with a self - developed real - time operating system (RTOS), and used a powerful noise - reduction algorithm to deal with complex sound fields. This courage to "start from scratch" ultimately increased the industry's general battery - life benchmark from 2 hours to 8 hours, and with weekly agile iteration, it established an efficient evolution paradigm for AR wearable devices.