Leading OEM/ODM Services for Cutting-Edge AI & AR Hardware Innovations
On January 13, 2025, VR陀螺 (VR陀螺) , represented by its General Manager and Editor-in-Chief, An Shanzi, held a live deep - connection session with Fan Qin, Deputy General Manager of博士眼镜 (Dr. Glasses) , Li Hongwei, CEO of雷鸟创新 (LeBird Innovation) , and Shi Qing, CEO of亿境虚拟 (Yijing Virtual) . The live broadcast, themed 【Traditional Meets Technology: Where Are the Surging AI Glasses Headed?】, delved into the current AI glasses market's potential opportunities, implementation challenges, and future development trends based on their experiences at CES 2025.
Fan Qin :
Participating in CES for the first time was incredibly exciting. CES serves as a window to the future of technology, showcasing its tremendous advancements. Beyond the familiar applications of AI in smart glasses, I also witnessed its integration into pet companionship, healthcare, and education. This experience truly highlighted how AI has transitioned from a theoretical concept to a part of our daily lives.
There have been numerous discussions about whether AI will rapidly replace many aspects of human life and cause widespread unemployment. I observed that while AI does make life more convenient, its complete substitution for humans in certain fields, such as ours, may still take time. Moreover, the significant presence of Chinese enterprises at CES, with over 1,500 out of 4,000 total companies, filled me with pride. It also underscored the intense competition in the Chinese market and the challenges faced by Chinese companies in expanding overseas.
During the exhibition, I first visited LeBird's booth and experienced the newly launched V3 product on January 7. It felt very lightweight and suitable for all - day wear. I also paid attention to other companies like Rokid, XREAL, and Li Weike, among dozens of AI/AR firms. The entire track is indeed very hot, and with the entry of capital and technological progress, the AI glasses industry is expected to accelerate its development.
Li Hongwei :
Having attended CES multiple times, I'd like to share several trends I observed. First, compared to last year's CES, there was a significant change: while last year there were many AI hardware products that garnered attention but had underwhelming actual performance, this year's AI - enabled products have become much more mature and have achieved a qualitative leap, making them acceptable to the general public.
The AI hardware showcased, including AI glasses, AI pets, and AI robots, have indeed gained acceptance from the mass users and have a high daily usage rate. Many companion robots have started to implement a subscription - based model, which is a notable change. Observing the evolution of technology has always been fascinating to me and has filled me with confidence that AI will be widely applied and become a product that users continue to use in the future.
Second, as previously mentioned, we have observed a continuous increase in the penetration rate of AI at CES. Although it may not be 100%, it is close to that, and I believe this is a positive trend. CES is not just a consumer electronics show; it also has the driving force behind the development of consumer electronics. In recent years, AI technology has made significant progress, and we have seen how it is applied to hardware and serves users, changing lives, which is a very positive aspect.
Third, being in the glasses industry, I have indeed seen the AI glasses market booming. No matter where you go, if you walk another 50 meters, you can see teams working on AI glasses, whether familiar or not. I believe that AI glasses are a crucial link in driving the future development of AI and the entire industry, and their role may be second only to AI itself. This is because AI relies on collecting a large amount of high - quality real - world data, and glasses happen to be an ideal product for collecting such data.
For example, when using LeBird V3, if you pick up a water bottle to drink, the AI glasses not only collect static information about the table and the water cup but also observe data about our actions with intent and interaction, such as picking up the cup. This data is extremely important for the development of AI, especially in the field of robotics. If robots want to understand the real world, they need a large - scale model with multi - modal, first - person perspective capabilities, similar to glasses. Moreover, robots also need to have the ability to interact with the real world and take proactive actions. The current proliferation of AI glasses is not only beneficial to the AI glasses industry itself but will also greatly promote the development of the entire AI industry in the long run.
Shi Qing :
At this year's CES, I noticed a significant trend: an increasing number of AI products are being implemented with high quality. Taking AI toys as an example, whether it's the touch feel or the interactive conversations, some products have performed exceptionally well, providing users with rich emotional value.
Focusing on our industry, AI glasses, as the important interface connecting the physical world with the AI backend, are undoubtedly the best hardware carrier. Compared with previous CES events, AI glasses have become the focus of attention this year, while the热度 (heat) of MR products has relatively decreased, with no particularly outstanding highlights. Although there are still opportunities for head - mounted displays in education, gaming, and movie presentations, the industry focus has clearly shifted towards AR and AI.
Regarding AI glasses, Yijing currently released AI glasses do not have a display function, but we will also consider developing a version with a display in the future. After all, AR glasses have a faster response time, enabling people to more quickly obtain AI feedback.
Case Shanzi :
Currently, the market's glasses products are mainly divided into the following versions - pure audio glasses, glasses with integrated audio and camera for shooting, and glasses with a display. What is the relationship between these three products in your view?
Li Hongwei :
In the long run, I believe this is a transitional relationship. There are several clear elements here. First, do we really need a display? I think the display has a strong intuitive nature and is key to human - computer interaction. Steve Jobs of Apple had a core principle in product design: whether the product conforms to the user's intuition. For example, in the case of AI glasses without a display, when we are in a shopping mall navigation and see several restaurants, it is difficult to distinguish which information the voice command outputs without the assistance of a screen. Therefore, displaying information is very intuitive and important for users. With the development of glasses technology, especially the progress of display technology and the integration of virtual reality, this will be a deterministic development trend.
The second question is about substitutability and transition. The key lies in whether AR glasses can serve as a general - purpose computing platform. Taking mobile phones as an example, although there are various devices on the market for taking photos and listening to music, such as cameras, music players, watches, etc., these markets are relatively small but still exist. Therefore, general - purpose computing platforms often have good compatibility, especially for these small - scene compatibilities. They have a richer range of application scenarios and contain more powerful potential energy. Based on these two reasons, I believe that some products without a display are transitional.
The third point is, of course, a technological issue, or rather a TPMP (Technology Product Market Fit) issue. With the development of technology, the cost gap between glasses with and without displays will narrow, including wearing costs and other related costs, and this effect will become apparent at that time. Additionally, I would like to talk about my view on timing. I am optimistic about the future and expect that AR glasses with displays will experience an outbreak in 2026. By 2028, AR will have a better display and a good blend of virtual and real effects.
Regarding the market size of these products, I believe that AI glasses with and without displays (including AI shooting glasses and AI + AR glasses) will both experience vigorous development in the next two years. We estimate that the global shipment volume may reach several million this year, and it is expected to reach the tens of millions next year. Once there is an overwhelming product design, its growth rate will be very fast. Among them, AR glasses with displays may be a bit later, but we judge that after 2026, they will also usher in rapid outbreaks. Perhaps in 2026, this type will only have a volume of several hundred thousand or a few million, but it will soon experience a multiple or even tenfold increase and maintain a development of doubling or higher multiples.
Shi Qing :
I also agree that AI glasses without display functions will exist in the long term because these products are lightweight and portable, and have obvious advantages in terms of lightness and comfort. However, when it will reach the market equilibrium point and the specific trend of change are still difficult to accurately predict at present. In addition, glasses with cameras but without displays will seize the market share of pure Bluetooth devices. Because the transformation of sound input/output and image input functions will be faster, and the pressure faced in the process of balancing lightness and digitalization of products is relatively small.
However, in the field of displays, we expect to wait for another 18 to 24 months before the relevant technologies gradually become mainstream. On the one hand, cost and power consumption are factors that must be considered; on the other hand, application scenarios and user experience are also extremely important. After all, there will be an obstruction in the middle of the field of view of glasses with displays, and users need some time to adapt. Of course, with the continuous popularization of cameras and displays, once they reach a certain level of penetration, I predict that the market for audio glasses or glasses with only cameras may be gradually eroded. Because for users who often wear glasses when going out, the price difference between these products may only be in the hundreds of dollars. Glasses with displays have richer functions and can take into account the wearing experience, and they are likely not to tend to choose low - configuration glasses with single functions.
Case Shanzi :
Due to the certain gap between the tech field and the traditional field, there may be differences in perspectives. What do you think are the key factors that ordinary users consider when choosing glasses?
Fan Qin :
The capacity of the glasses market in China is about 100 billion. In fact, before the emergence of the phenomenon - level product Ray - Ban Meta, the attention of the market to the glasses industry was not high. The AI glasses trend is coming, and as a glasses practitioner, I am very excited because the appearance of smart glasses provides us with a new opportunity and increases the market's attention to the entire industry, bringing new increments to the glasses industry.
In my personal understanding, the development of smart glasses is not only the progress of intelligent hardware, including the continuous miniaturization and lightening of modules. At the same time, with the integration of AI large models, the wave of glasses intelligence is coming. This has made smart glasses evolve from being too heavy to wear for a long time to being able to be worn all day, and the styles are now close to the level that most of us can accept. The weight of ordinary glasses is mostly between 30 - 40 grams, which consumers can accept. The current weight of smart glasses is about 50 grams, and LeBird V3 has achieved 39 grams, with very positive consumer feedback.
Another point is that the needs of different consumers are unique. Some users believe that the combination of glasses audio and AI interaction can already meet some needs, and some users have unique demands for shooting. In addition to the audio and video functions mentioned earlier, I think AI glasses may appear in various scenarios in the future, such as applications in medical care, education, and sports. Of course, if a glasses needs to integrate more functions, the requirements for it will be higher.
Case Shanzi :
Can you talk about the equipment replacement cycle, price, application scenarios, and SKU of ordinary glasses market?
Fan Qin :
In China, consumers usually purchase glasses with a cycle of one to one and a half years. However, today's consumers are different from the past. They often have multiple pairs of glasses to adapt to different scenarios, such as sports, office work, and travel. In addition, there are many categories of glasses. Besides the very rich styles of frames to meet consumers' pursuit of aesthetics and fashion, the functions of lenses are also diverse, such as children's defocus lenses, adults' fatigue - relieving glasses, elderly people's all - in - one reading glasses, and myopic sunglasses. In fact, everyone has more than one pair of glasses, and if we add a pair of smart glasses later, I think the application scenarios will be even wider.
The SKU quantity of a store is usually between 800 and 1,500, depending on the size of the store, and consumers have a high degree of choice when entering the store. At present, the newly launched smart glasses are still in the initial stage, with fewer styles. However, manufacturers have made many improvements in hardware design, such as some supporting replaceable temples, and I see that the overall hardware design is also constantly progressing and innovating.
In fact, everyone's needs are different, and each pair of glasses is a unique customized product. From the frame to the lenses, diopter, the height of the nose pads, the length and tilt angle of the temples, and other details, these differences make the customization of glasses complex. Moreover, glasses users will also consider personalized needs, making it difficult to be like other standard - form consumer electronics that conform to the "80/20 rule." If ten people go to eat, and eight of them wear similar glasses, they are probably not willing to accept it.
Regarding the price,博士眼镜 (Dr. Glasses) has several brands. The price of presidential glasses is often above 4,000 yuan, the price of the博士眼镜 (Dr. Glasses) system is about 800 - 900 yuan, and the price of the石人工 (Shi Ren Gong) system is around 300 - 400 yuan. We also have various promotional activities in our online flagship store, with prices ranging from over 100 to over 300 yuan, depending on the user's needs. Given the huge Chinese market and the extremely high incidence of myopia, especially in first - tier cities, where the myopia rate among college students is close to 90%, such a large group actually has a huge demand. Therefore, I don't think we should worry too much about the price issue. This time, the 1,799 - yuan LeBird V3 jointly launched by us and LeBird has also been widely welcomed by the market.
As for the average number of glasses per person, most vision - care consumers have more than two pairs of glasses, especially for those with a diopter of more than 300 degrees. Glasses have almost become a part of their body. They can't live without glasses except when sleeping, and it's very common to have spare glasses.
Case Shanzi :
What solutions do AI glasses have for the current myopic population?
Li Hongwei :
Since our end - user products are directly面向消费者 (面向消费者 - facing consumers) , and we have considered this factor in the early stages of product development. For example, for glasses without displays, users just need to put in the lenses. As mentioned by General Fan, we have an in - depth cooperation relationship with博士眼镜 (Dr. Glasses) . Users can provide their myopia diopter, and we may embed the myopia lenses when shipping; they can also choose to have their eyesight tested and fitted offline at博士眼镜 (Dr. Glasses) stores, which mainly depends on the user's choice.
For products with displays, LeBird has made different deployments in its medium - and long - term strategies. In the short term, we will focus on optimizing the ID, such as combining the lenses and waveguides to ensure that the fit is good enough so that users can't tell that it's composed of two lenses. In the long run, we are also exploring some very promising new technologies aimed at solving the so - called three - in - one problem at the technical level, that is, the integration of myopia lenses and waveguide lenses. In addition, we are also discussing some promising waveguide technologies that can achieve instantaneous adjustment of refractive power, but this is still at the conceptual stage.
Case Shanzi :
How about the growth of AI glasses business this year?
Shi Qing :
This reminds me of the period when tablet computers emerged in 2010. There were many products on the market with a wide range of applications. In fact, manufacturers of computers, accessories, the Internet, AI, and XR have all shown interest in this area. It can be预见 (预见 - foreseen) that this will be a stage of vigorous development. However, I believe that not every company can successfully develop such products because it involves everyone's aesthetic differences and multiple considerations including safety.
In the exploration of AI glasses, confidence, determination, and execution are crucial. Teams represented by General Hongwei have particularly strong execution and are at the forefront of rapid development, setting off in the right direction earlier, which enables them to achieve success quickly. However, due to the wide range of scenarios, I believe that other companies also have many opportunities in their respective advantageous细分赛道 (细分赛道 - niche tracks) .
Fan Qin :
We started paying attention to and researching smart glasses in 2021, and began selling smart glasses in 2023, having successively cooperated with some well - known brands. Many consumers have gradually learned more about AI glasses by walking into our stores.博士眼镜 (Dr. Glasses) has always been the "last mile" of smart glasses, providing consumers with the best fitting and experience scenarios. They may have heard about it online before, but when they truly experience these smart glasses, especially those with AI and waveguide display functions, many consumers are still surprised and willing to buy and try.
Another point is that I have learned that some people have started to give AI glasses as gifts. Especially now that the Spring Festival is approaching, when everyone is considering what to give to their relatives and friends, consumers are not lacking in basic items, but receiving smart glasses as gifts, many people are willing to accept them, and there are also many people who order several at a time. I think the market for smart glasses as gifts is also broad, not limited to personal use.
Case Shanzi :
How do you view the demand for multiple SKUs of smart glasses?
Li Hongwei :
Not only glasses, but many categories in this era actually have multiple SKUs. The clothing industry is a typical example, and we have witnessed the market development process of fashion, fast fashion, ultra - fast fashion, and super - super - fast fashion. However, electronic consumer products follow a different logic. I think combining the two may produce interesting results. LeBird is also trying such a combination because the disruptive nature of new industries or new directions is essentially the disruption of the entire value network. From this perspective, it seems to be a cost, but on the other hand, it may be a potential opportunity to disrupt the smartphone industry.
When designing SKUs, it is necessary to consider its fashion attributes as a decorative item. We have already carried out preliminary analysis in the design of glasses. In addition, we have considered some basic methods to reduce the costs caused by multiple SKUs. Although adding cameras and displays in front increases complexity, there are still technical means to achieve this goal. LeBird has actually done a lot of preparatory work in this process, and may subsequently display the final results or release related products to everyone.
Another point I want to add is that dealing with multiple SKUs is also a problem that needs to be balanced. Traditional glasses and electronic consumer products each have their own characteristics, and the final result of AI glasses is likely to be in a middle state between the two. It is neither as numerous in SKUs as traditional glasses nor as few as electronic products. Product companies themselves need to have the ability to judge what kind of design is more likely to attract current target users. The demand for AI glasses is not that great yet, and the volume is not that much. But according to the results of our user research, there are still traces to follow regarding what kind of SKUs users like or what kind of ID designs they prefer.
Shi Qing :
The demand for multiple SKUs of AI glasses needs to consider the cost issue in the early stage. Taking a well - made mold as an example, its cost is usually between 400,000 and 1 million, depending on the required precision. If the brand positioning is higher, the cost may be even higher. Therefore, the customization fee that is apportioned to each product must be considered. Components such as batteries, speakers, etc., also need to be customized according to the product's characteristics, all of which will increase the cost.
Moreover, on the software side, including wake - up words, optimized algorithms, ESI, camera shake, and night noise reduction algorithms, etc., may all affect the final cost. Overall, if the sales volume within a year is less than 50K, it will be very depressing. However, as we mentioned earlier, for traditional glasses, a single style reaching 50K is already a相当不错的成绩 (相当不错的成绩 - pretty good achievement) .
Case Shanzi :
When facing customers, have you encountered some relatively special needs, after all, glasses are relatively personalized products?
Shi Qing :
We are committed to finding the greatest common divisor and focusing on the commonalities. In fact, for the lenses of smart glasses, you will find that the changes are not great. We mainly make some improvements while maintaining the commonalities. The various changes and differences in the front frame can be handled by our partners. We focus on the optimization of the lens components, including batteries, speakers, motherboards, and high - density wiring, etc. By improving the process of wiring with more bends, we reduce costs and increase the yield rate to match the changes in front, thereby ensuring the stability of the product's launch time and other aspects. This may be the first step in optimizing the production process.
After everyone has gone through 12 or 18 months of in - depth thinking and exploration, I believe people will focus more on their own areas of expertise. I fully agree with General Fan's earlier view on diversified operations. In fact, I heard that in the traditional glasses industry, a scale of 50,000 units has already been a very large volume, which is a challenge for some big companies, because it involves issues such as turnover, inventory, and yield rate.
As for the functions, I think AI glasses are not in a hurry to achieve the maximization of functions. The first step is to design glasses that are beautiful, fashionable, lightweight, and comfortable. This is crucial. Subsequently, we can gradually add more functions, allowing everyone to gradually become familiar with and accept this category, instead of providing too much information at once. We have ventured into the MR market and found a major problem: the speed at which ordinary consumers accept new things is far lower than our expectations, and overly complex products are not accepted by them.
I think that in the promotion process, everyone should try to gradually increase the complexity in product definition and other aspects, giving customers an adaptation process. We cannot create demands just because customers want all functions. Sometimes everyone is too eager to pursue the unique advantages of their own fields, but we should also avoid creating unnecessary demands.
Additionally, I would like to add that recently Yijing jointly established an AI glasses alliance with博士眼镜 (Dr. Glasses) . The original intention of this alliance is that the member units within the alliance can form a supply chain cluster like the Shenzhen model in the past. If this industry is mature enough and everyone can well organize their own areas of expertise, I firmly believe that it will be able to provide various efficient, fast, and low - cost solutions for the global market, and it is possible to create products that can compete with Meta and other big companies. We also hope that more strong member units will join in.
The AI glasses roundtable discussion provided valuable insights into the current state and future prospects of the AI glasses market. The participants highlighted the importance of technological innovation, user - centered design, and strategic partnerships in driving the development of AI glasses. As the industry continues to evolve, it is clear that AI glasses have the potential to become a transformative force in the way we interact with technology and the world around us.