2025 has been a year of explosive growth for AI applications, and smart glasses, regarded as "one of the best carriers for the implementation of large AI models," are undoubtedly one of the most bustling tracks in the tech industry. From Meta's rapid sales growth of Ray-Ban Meta to the sudden heating up of the domestic "Hundred Glasses Battle," nearly everyone from startups to internet giants and smartphone manufacturers is talking about AI/AR glasses. The rapid influx of capital has quickly propelled this once-quiet industry into the spotlight. However, behind the "Hundred Glasses Battle," few products have truly achieved sustained shipments and user retention. Shipment volumes, channel capabilities, and technological accumulation are becoming the watershed, with capital and resources increasingly concentrating towards a handful of leading companies. At such a juncture, on the second working day of 2026, leading AR glasses manufacturer LeiNiao Innovation announced the completion of a new funding round, jointly invested by funds under China Mobile and China Unicom. This also marks the first time that mainstream Chinese operators have jointly invested in the smart glasses sector. "Over the past year, significantly more investors have rushed into the smart glasses industry," revealed LeiNiao Innovation founder and CEO Li Hongwei in a recent exclusive interview, adding that this round was heavily oversubscribed, reaching two to three times the originally planned amount. Simultaneously, he pointed out that although investor enthusiasm is currently high, funds are still converging towards the top players. Counterpoint data shows that LeiNiao Innovation captured 24% of the global market share in Q3 2025, ranking first in the global AR smart glasses market for two consecutive quarters. Furthermore, it is understood that last year, LeiNiao Innovation's overseas revenue accounted for over 30% of its total, with expectations to exceed 50% in 2026. Globalization is a key strategy for LeiNiao's development, with North America, Europe, Japan and South Korea, Australia, the Middle East, and Southeast Asia being its current and future key markets. Regarding the reasons why mainstream Chinese operators chose to jointly bet on the smart glasses track at this time, Li Hongwei believes the industry's development trajectory is now very clear; everyone sees the sector growing rapidly and products maturing quickly. "The operators judge that the industry has reached an inflection point for explosive growth." From the perspective of the operators' own development, investing in LeiNiao Innovation is primarily based on two considerations: first, a strategic judgment that operators aim to position themselves in the "next-generation phone" arena; second, the belief that now is the right time to initiate cooperation with smart glasses manufacturers. It is worth noting that alongside the cash investment from funds under China Mobile and China Unicom, significant breakthroughs in products and channels are expected for LeiNiao Innovation – during CES 2026, LeiNiao will showcase its first eSIM AR glasses, which feature a binocular full-color display solution and a built-in eSIM module. "Since smart glasses are the next generation of phones, they need to be relatively independent," Li Hongwei stated, noting that whether it's full-color MicroLED or eSIM, it's evident that only LeiNiao is genuinely working on the "next-generation phone." The business cooperation between LeiNiao Innovation, China Mobile, and China Unicom is systematic. On one hand, they will co-develop products; besides eSIM, operators can also assist LeiNiao in deploying edge computing capabilities, addressing the weaknesses of smart glasses in computing power and latency. On the other hand, the operators' vast user base and channel capabilities provide a practical path for scaling LeiNiao's products. "Over a decade ago, operators facilitated the rapid popularization of smartphones through schemes like 'get a phone with prepaid credit,' which spread quite quickly," Li Hongwei remarked, indicating that LeiNiao Innovation will also collaborate with operators on similar 'prepaid credit for glasses' promotions. Once this model matures, it could become a significant inflection point for industry-wide adoption. Undoubtedly, the capital betting on smart glasses represents a contest for the "next-generation mobile terminal" and the "AI super gateway." The future potential of smart glasses is a consensus within the industry; the debate lies in when, and who, will bring smart glasses into millions of households. "I don't quite agree with the view that 2025 is the first year of explosion or the breakthrough year for smart glasses," Li Hongwei expressed, arguing that the past year did not see fundamental breakthroughs in product capability. The main characteristic was industry buzz, with more people recognizing the market opportunity, but attention to experience standards remained somewhat lacking. Li Hongwei stated that 2026 will be the true first year for smart glasses, with 2027-2028 being the "iPhone moment." There is a distinction: 2026 involves "products that deliver value along the way," meeting user experience standards, while 2027-2028 will achieve a state akin to the original iPhone. This assessment contrasts with much of the market's optimism. Over the past year, smart glasses frequently took center stage, with some features gaining mainstream attention, significantly boosting industry visibility and leading many to label 2025 as the "first year" or "breakthrough year." Clearly, in Li Hongwei's view, buzz does not equate to maturity. Many smart glasses products resemble mere feature permutations, still far from being devices ordinary users are willing to wear all day and use long-term. This also explains why many products gain short-term attention but struggle to build lasting competitiveness. Li Hongwei divides the upcoming development of the smart glasses industry into three stages: The first stage, from now until the "iPhone moment," will see numerous players, with the core question being who can truly build the next iPhone. The second stage, from the iPhone moment to the "iPhone 4 moment," is the process of a revolutionary product becoming mainstream. The third stage, from the iPhone 4 moment to determining the winners, involves surviving players competing comprehensively across all dimensions. On the eve of the industry's explosion, major players must go all out to become the first company to create the "iPhone." Li Hongwei emphasized that the core logic of this phase is identifying who has the determination to build the next iPhone and possesses sufficient resources, including capital and talent, to move fast enough. It is understood that the latest cash funding round for LeiNiao Innovation exceeded 1 billion RMB. Li Hongwei revealed that the core use of these funds is research and development, aiming to quickly bring 2026 products up to experience standards and expand advantages in key technologies like optical display, spatial computing, and AI. "Many products were launched in the industry last year, but frankly, almost none truly met user experience standards," he said, noting that whether users find the product excellent and are willing to recommend it to others after use is his core criterion for judging product maturity. However, Li Hongwei believes that now is the time when sustained effort in both new technologies and past engineering investments can finally achieve these experience standards. Returning to the products themselves, the distinction between AR glasses and AI glasses is not entirely unified externally. Their technical routes, functional positioning, and marketing often overlap or are ambiguous. Some manufacturers view AI glasses as a transitional form towards AR glasses, while others frequently define their products as "AR+AI" glasses. Li Hongwei believes the future smart glasses track will mainly comprise the following product forms: First, AI glasses without displays, a category closer to the logic of products previously launched by some major players, whose primary strategic value is data collection and enhancing existing AI ecosystems on phones. Second, AR glasses with displays, subdivided into two categories: one resembles normal glasses with very few features, and the other offers relatively powerful functions but comes with higher weight and cost. On the technical path, LeiNiao has chosen a relatively "heavy" route: full-color MicroLED waveguide display, low-power system architecture, and deep integration of AI with spatial computing. This is a undoubtedly challenging but correct path, seen by Li Hongwei as the necessary route to the "next-generation terminal." On the first working day of 2025, LeiNiao Innovation also dropped a bombshell on the smart glasses industry by announcing a strategic partnership with Alibaba Cloud. Alibaba's Tongyi series of large models will provide exclusively customized, full-suite AI technology and product support for LeiNiao's AI glasses. At that time, Li Hongwei revealed that exploring Killer Apps would be the top priority in the collaboration with Alibaba, with on-device large models ranking second or third. Now, Li Hongwei states that creating killer apps remains LeiNiao Innovation's most important task, and they are now at a stage of being ready and nearing launch. On the long and challenging track of smart glasses, the current buzz is just the beginning. The true watershed likely still lies ahead. Following is the full edited transcript of the interview. The smart glasses industry is reaching an inflection point for explosive growth. First, it's a process of mutual attraction because everyone has reached the same conclusion about the industry's prospects. From the perspective of the investing operators, they judge that the industry has reached an inflection point for explosion. Although the industry's iPhone moment hasn't arrived yet – I believe that will be in 2027 or 2028 – the current situation is very clear. They see the industry growing rapidly and products maturing quickly. For the operators, there are two considerations. First is strategic consideration: what is the next generation of phones? They want to position themselves in the next-generation track, which aligns more with their capital-level considerations. Second is business cooperation consideration: from a collaboration standpoint, it's about when the products are mature enough to genuinely engage in business cooperation. Because they serve mass users – each operator serves hundreds of millions, even billions – they want to offer a good product experience. These are the two key judgments, and they both feel the timing is right. Frankly, I don't quite agree with that view. I don't think 2025 is the first year or breakthrough year for AI glasses. When will it be? Probably 2026, because products will truly exceed user experience standards. The change in 2025 was that many people started working on this, but the important question is whether there was a real breakthrough. What's the underlying logic? I think the logic for many entrants is that they see the market opportunity but haven't yet grasped the market's essence. Even when people talk about it being the first year, they miss this essence. If we must call it a first year, I believe 2026 is the first year, and 2027-2028 are the iPhone moments. There's a distinction. People focus more on the surface-level buzz rather than the underlying changes. There's very little discussion, even within the industry, about the "experience standard." Many current products are just interesting combinations of features, but the real issue is whether you've made a product that meets the experience standard. Some become pessimistic because of this, saying the direction isn't worth pursuing. I think this also misses the essence. They don't see that the experience standard will change in 2026. I divide the industry into three stages: Stage one is from now until the iPhone moment; all sorts of players will enter. The core is whether you can actually build the iPhone. Stage two is from the iPhone moment to the iPhone 4 moment; major players will formally enter the fray, and the core remains product leadership. What does formal entry mean? Does Xiaomi count as formally entering now? I don't think so, judging by the product outcomes. They haven't deployed their main forces and core resources; if they had, the results should be better than what we see. Stage three involves the remaining players engaging in fierce, head-to-head competition across all dimensions. The core logic of the first stage is identifying who has the determination to build the next iPhone and possesses sufficient resources, including money and people, to move fast enough to achieve it. From a strategic perspective, they are betting on who will ultimately succeed in this track, and their judgment is that LeiNiao has the highest probability of success. From a product perspective, they favor LeiNiao's current offering as the best product on the market. For the operators, several characteristics of LeiNiao's product indicate its quality. First, we have a very large user base. Recent Q3 data from Counterpoint shows we rank first globally in the AR glasses space with a 24% market share; one out of every four AR glasses sold globally is from LeiNiao. A large user number signifies the product has been validated. Second, the product experience standard is good, with positive user feedback. Third, LeiNiao's products are in an appropriate price range for the category. For them, just as it was "prepaid credit for phones" before, now it can be "prepaid credit for glasses." This requires the glasses to be at a suitable price point. "Prepaid credit for glasses" could mark the beginning of an inflection point. So, the product stands out on these dimensions. As for how they judge LeiNiao as the company with the best chance of success? I think there are several aspects: First, compared to other companies, LeiNiao's most important advantage, relative to the entire industry, is its leading accumulation of product technology. For a new category, you must always provide valuable products to users, which hinges on product and technological leadership – something they recognize, especially leadership on the correct path. Currently, it seems only Meta and LeiNiao are genuinely working on the "next-generation phone," meaning full-color display AR glasses as standalone devices. Another point is that they can only invest in startups, and among startups, LeiNiao is increasingly leading due to a Matthew effect. From a financing perspective, one can also see focus converging on the most promising top-tier companies; they see this too. Of course, behind this are the team, strategy, and a series of other factors. There will be systematic business cooperation, mainly in two aspects: First, co-developing products. For example, we are adding eSIM to LeiNiao glasses. Previously, glasses had to pair with phones, and the experience wasn't great – pairing via Bluetooth. Strategically, this method creates perpetual dependence on the phone. But we know glasses are the next-generation phone, so they need relative independence. That's why I said only LeiNiao is genuinely making the next-generation phone, whether it's full-color MicroLED or eSIM. Beyond eSIM, there are many other collaborations. For instance, AI is crucial now. Through our cooperation with operators, we can deploy not just in the cloud but also at the edge. Operators have integrated significant AI computing power into base stations, enabling faster AI processing. Relying solely on the cloud introduces network latency issues. Another example is user video/photo storage in cloud albums – there are many potential business collaborations with operators. So that's the first point, product-level cooperation. The second aspect is user promotion. Over a decade ago, operators facilitated the wave-like popularization of smartphones through "prepaid credit for phone" deals, which spread very quickly. So, in the future, when we launch "prepaid credit for glasses" promotions, the adoption speed could be rapid. This has two benefits: it covers a vast user base, and operators maintain trust within their user systems; the model of getting something with prepaid credit is already mature, which is valuable for us. We have a general idea of the price range for the glasses offered, but the specific timing isn't fully decided yet and isn't something we can disclose publicly. The most crucial focus is research and development. What's the problem with current industry products? We can be frank: many products were launched this year, but almost none met user experience standards. After using the product, can users say it's excellent, recommend it to others, and find it sufficiently mature? I don't think we're there yet. For LeiNiao, we have built industry-leading technology over the past four years, so we know the current state of the technology. I believe we've reached a point where sustained effort in both new technologies and past engineering investments can achieve the experience standard. The raised funds are truly intended for significant R&D investment to quickly bring our 2026 products up to that standard. First, the number of investors has increased significantly, so we were heavily oversubscribed. The amount we raised is 2 to 3 times what we initially planned, and we even turned down funding offers similar in size. So investor enthusiasm has grown considerably. Second, capital is concentrating towards the top players. Investor passion is high now, but if you look at who is raising substantial funds, it's mainly the leading companies in this track. This effect is quite evident because the industry's development has reached a stage where it's not about fragmentation; capital gathers around head companies before the iPhone moment. It might disperse slightly after the iPhone moment, but I believe the general trend is towards concentration. Because after the iPhone moment, confidence grows stronger. If you can't invest in the top player, you might invest in others. But at this current moment, you can still invest in leading startups if you want to. We have such plans, but there's no public timetable yet. Our overseas revenue占比 is over 30% now, and in 2026, overseas revenue will基本上 exceed 50%. LeiNiao is inherently a global company; that's very clear. Pursuing this direction (smart glasses) necessitates globalization. Moreover, I think this generation of entrepreneurs doesn't see China and overseas as two separate markets like before; now, everyone considers the globe as one single market, with a more globalized mindset. LeiNiao has a strategic deployment plan: we focus on one market at a time rather than spreading efforts thinly across many markets at once. We started with China, then entered the US, beginning focused efforts there in mid-2024. This Black Friday and Christmas season, you can see our significant presence in North America, Europe, Japan, and South Korea. So, if you look at the best-seller lists on Amazon in the US and Europe, you'll find LeiNiao products. That's our deployment strategy. Due to these channel deployments, we will have a strong presence in North America, Europe, Japan, South Korea, and a series of other overseas regions in 2026, so naturally, overseas revenue will surpass domestic revenue that year. If we can achieve leadership in China and the US, replicating that should lead to leadership elsewhere. If you have the capability to lead, you just need to enter more markets. Of course, products require some customization because user needs differ by country. For glasses, we need to make adjustments. The software ecosystem will also differ, and we have tailored marketing and branding strategies for each region. In our globalization approach, we didn't follow the traditional path of simply exporting. We relatively quickly jumped to localization. So you'll see our marketing and branding tactics differ across countries. For example, in China, we focus more on grassroots seeding ("planting grass"), while overseas, we emphasize brand building. For instance, we will launch IP-collaborated products in North America early in 2026, focusing more on branding. There are differences in channels too. We primarily operate based on two rules, the most important being the actual local channel landscape. China and overseas are different. China is now predominantly online for various electronics, while the US is still mainly offline, with an online-offline split around 70-30 or 60-40. Although we entered the US online first, we will relatively quickly move into offline channels like Walmart, Best Buy, Costco, Target, Sam's Club, etc. That's the main difference. We'll start with major retailers, leveraging existing mature systems more. We will have our own flagship stores in the future, but their short-term purpose will primarily be for brand building and experience. We are focusing on several key technical directions: First is the optical display part, so many of our glasses next year will have displays. Core issues in optical display include: 1) Imperceptible Display: Current MicroLED waveguide glasses still have issues. For example, wearing monochrome glasses tints your view of the world yellow, and light transmittance isn't high enough. For someone like me who doesn't need prescription glasses, I feel it obstructs my view; there are also rainbow artifacts. And the person opposite me might see light leakage from my glasses, an abnormal patch. That's the first point in display – imperceptibility. 2) Performance issues, which include power consumption. How do we minimize power consumption so the glasses can be worn all day? There are many challenges here, including display brightness, resolution, and other metrics. Second is AI. While everyone says AI is one of the most important application scenarios for glasses, often it's just about putting AI on glasses. We've done some optimizations, like our exclusive partnership with Tongyi to speed it up and improve effectiveness, but it's not enough; there hasn't been a qualitative leap. So we hope to achieve qualitative breakthroughs – creating killer applications based on AI on the glasses. Past applications are mostly about asking questions, not much different from using AI on a phone – telling you the weather or what flower that is. There are no killer apps yet. We are training our core scenarios based on large models, such as long-term Memory, to develop them into killer applications. Third is spatial computing algorithms. People haven't used spatial computing capabilities much before. For example, when translating, you look at a street sign while walking; can the translation result be overlaid on it? We are working on that. We're not just developing the algorithms themselves; the papers we publish are among the core papers in this field. We are also integrating algorithms with chips, optimizing device performance at the production end. Fourth is optimizing basic experiences, like hardware performance, software performance, battery energy density, glasses size, weight, etc. I'm breaking it down from a technical perspective. From a user's perspective, they will see glasses comfortable enough to wear and AI killer applications, with a high experience standard. No, it remains the most important thing. We are basically at the stage of being ready and nearing launch. We've done a lot of underlying technical work. I agree. It's a native AI device, free from the historical baggage of the phone ecosystem, and can add value during times users aren't looking at a screen. Looking back at history, we always think frontal assaults are good, but they usually aren't; success often comes from flanking attacks, disruptive successes. I've always said glasses are the best carrier for AI. Besides seeing what you see and hearing what you hear, another point is that they are native AI devices. You don't have to compete within those existing ecosystems; that competition takes a long time to resolve, filled with entrenched interests. But as a native AI device, smart glasses have no such baggage and can excel. Yes. It's not entirely a computing power issue on glasses; it's a power consumption issue. I think it will take some time to bridge the gap because glasses need to be comfortable to wear; they can't have very large batteries. So for AI on glasses, it's an end-edge-cloud collaborative system. It doesn't necessarily have to be solely on the device, though some things can be placed there. Current chips on glasses can run 1b parameter systems, while phones can run 6-7b systems. It's more about power consumption. First, the Agent ecosystem on glasses is different from that on phones. Glasses will likely start with an ecosystem different from phones; some have summarized it this way, and I agree. Typically, we look at phone screens for 6-7 hours a day. Besides 8 hours of sleep, there are another 8 hours when we aren't looking at a phone screen. During that time, we also want AI empowerment, so that's a good territory for glasses to发挥作用 first. I might not compete with you head-on entirely; I first find a very effective niche. So its ecosystem isn't completely identical to the phone's. When glasses need to interact with the phone ecosystem, at least there hasn't been a situation like apps blocking Doubao's phone assistant yet, because glasses haven't attacked the phone's core territory. Many are handling it well now; many have opened up Agent system interfaces, like Alipay's "Zhi Xiaobao." So the mindset towards glasses is more open. It also doesn't fully cover the entire phone ecosystem yet. I think progress is relatively slow now. The essence of ecosystem building isn't just building an ecosystem; it's creating a sufficiently good product first. If the product is excellent, software vendors will be willing to join proactively. Many claim they are constantly building ecosystems, but I think that's misguided. You create an iPhone; when the first-generation iPhone launched, there was no App Store, yet people were clamoring to develop for it. Yes, once smart glasses truly reach their iPhone moment, many software vendors will naturally join. Trying to force it now is mostly futile. The only benefit now is that you *must* find a few killer applications. During that process, cooperation with the ecosystem is necessary. But building a broad ecosystem now – what's the use of having 10,000 apps on glasses if nobody uses the glasses? It's meaningless. We are working on this now, just haven't released it yet. But for this Agent, the primary method we are using currently relies on API interfaces, not GUI Agent methods. GUI Agent methods are less efficient, a last-resort approach. We have GUI Agent capabilities too, but we prioritize using API methods. We expect our new product launch to have that effect; it should be released in the first half of 2026. People have said "inspiring" before, but the final effect didn't meet expectations. This time, I feel more confident.
Comments