Yang Longsheng, CEO of Yingmu: It's a good thing that Xiaomi has entered the AI glasses market. Everyone's sales have tripled or quadrupled.
Author: Qiu Xiaofen
Editor: Su Jianxun
In 2025, the popularity of AI glasses is obvious to all. It's hard to imagine that until last year, only a few state - owned enterprises showed interest in the glasses market.
"In 2025, strategic investors from various industries and market - oriented funds have all flocked in," recalled Yang Longsheng, the CEO of INMO Technology, one of the so - called "Four AR Dragons" in the industry. He remembered that their recent round of financing was completed within just two months.
After meeting with over a hundred investors, they finally secured 150 million yuan in Series B2 financing jointly invested by Puhua Capital, Liangxi Industrial Development Group, and Shenqi Capital (under 58.com).
Having been in the industry for many years, starting from being the head of Coolpad's smart wearable department to founding his own company, Yang Longsheng is accustomed to the industry's ups and downs.
INMO AIR 3. Picture from the official source.
When OpenAI and Deepseek led the wave of large - model development, it also ignited the industry's enthusiasm for exploring how to implement AI. AI glasses have naturally become one of the answers - glasses are inherently closer to the human eyes, which are used for information acquisition, and they also have a wide user base, making them an excellent scenario and entry point.
This wave of enthusiasm truly began in 2024 when Meta collaborated with the veteran glasses manufacturer Ray - Ban on AI glasses, which became a huge hit. The sales of this product exceeded 2 million units in 2024. After the success of the Ray - Ban Meta glasses, at the end of last year, domestic tech giants and startups took over the baton.
Before Xiaomi, INMO launched its AI glasses, INMO AIR 3, in November last year. However, different from most current AI glasses products, AR remains the company's main focus. INMO AIR 3 is an all - in - one AR glasses that supports 1080P display and binocular full - color, and it also integrates AI semantic interaction, spatial computing, multi - screen collaboration and other capabilities.
INMO AIR 3. Picture from the official source.
The launch of Xiaomi's AI glasses at the end of June directly set off a new high. Yang Longsheng remembered that in 2025, all the investors who approached him asked the same three questions: "What's your view on the AI glasses field?", "What impact will Xiaomi's entry have on the industry?", and "How can startups resist the attack of tech giants?"
Regarding Xiaomi's entry, Yang Longsheng holds a more optimistic attitude, believing that it has opened a new chapter for AI glasses. Yang Longsheng told Intelligent Emergence that according to monitoring, since the end of June, the traffic of the smart glasses market has been ten times that of before, and the sales of each glasses manufacturer are expected to increase by two to three times this year.
With the gathering of tech giants, general users, and capital, the eve of the "Battle of a Hundred Glasses" has arrived.
Looking at the current situation in the market, there are three types of players in the AI glasses field -
One type is internet giants, such as Xiaomi and Baidu; another type is AR glasses manufacturers, with the main players being the "Four AR Dragons" (INMO, Xreal, Rokid, and Thunderbird). These players are expanding into the AI glasses field; the third type is new entrants without glasses - making experience who are seeking transformation, such as AnkerMake and a group of manufacturers in Huaqiangbei.
In China, it's very easy to make a "camera in the form of glasses". That's why so many players with different backgrounds are eyeing this field. Among the two hottest sub - fields of AI hardware, robotics and AI glasses, the latter has a much lower threshold. However, Yang Longsheng said that there are actually many "pitfalls" in the AI glasses market.
In the long run, Yang Longsheng believes that the competition in the future AI glasses market will focus on three aspects.
First, in terms of hardware, the key to the competition of AI glasses lies in whether they can achieve lightweight and miniaturization like ordinary glasses, which determines whether users can wear the product continuously. Additionally, he said that the ultimate development of AI glasses is inseparable from optical display, and it also tests how manufacturers can optimize the solution further (such as yield rate, cost, display effect, etc.).
In terms of the crucial AI capabilities, if manufacturers simply use general AI large models, the homogenization of AI glasses will be prominent. Yang Longsheng said that as one of the intermediate layers, AI connects hardware and scenarios, and it tests how manufacturers can make good use of AI to improve data acquisition and achieve spatial computing capabilities.
On the supply - chain level, it tests how AI glasses manufacturers can reduce hardware costs to accelerate their entry into the mass market.
INMO AIR 3. Picture from the official source.
At the beginning of the "Battle of a Hundred Glasses", there are still only a few enterprises that have truly achieved product maturity, user verification, and a clear business path. The "Battle of a Hundred Glasses" is actually a multi - dimensional competition. "Tech giants are currently in the trial - and - error stage, and there is still a window period of two to three years left for startups," Yang Longsheng said.
The following is the actual record of the exchange between 36Kr and Yang Longsheng (slightly edited according to the actual situation)
Xiaomi's Entry: Sales of Each Manufacturer Increased by Two to Three Times
Intelligent Emergence: Xiaomi's entry into the AI glasses field has attracted wide attention from the entire industry. What's the significance of this for the industry?
Yang Longsheng: Xiaomi's entry is a good thing for the whole industry. AI glasses used to be for a niche group, but now more people are starting to pay attention to this field, which has a very obvious driving effect on market education and public awareness.
In terms of specific traffic, during the launch period, Xiaomi's entry boosted the overall market traffic by nearly ten times. Judging from the daily average traffic, I estimate that it will drive the traffic of the whole industry by about three times.
For each manufacturer, the overall sales this year may increase by two to three times. Specifically for us, after Xiaomi's launch, our sales have increased by about 1.5 times.
Intelligent Emergence: So you don't think Xiaomi's entry will erode the market share of others, but instead hold an optimistic attitude?
Yang Longsheng: Currently, there are still obvious differences between startups and tech giants like Xiaomi in the smart glasses field. It's not yet the stage where they really erode each other's market share.
Intelligent Emergence: Can you break down the cost structure of a pair of smart glasses, taking Xiaomi's glasses as an example?
Yang Longsheng: Firstly, Xiaomi's glasses are the simplest form of AI glasses. The major part of its cost is the chip. Just the Qualcomm AR1 chip costs about 44 US dollars. Together with the components on the motherboard, it may cost six or seven hundred yuan.
Adding the peripheral sensors and optoelectronic circuits, it may cost eight or nine hundred yuan. With the structural parts like the shell, the overall BOM cost should be around 1,200 yuan. This is the cost structure of the simplest type of AI glasses currently.
Our glasses are of a higher - level form, with an added display unit module. We have well - controlled the BOM cost, which is basically similar to that of a photo - taking glasses. Most of our motherboards and sensors use domestic solutions.
In the whole industry, the cost of glasses with display and AI functions may reach 1,700 to 1,800 yuan.
Only State - Owned Enterprises Dared to Invest in Glasses Companies Before, Not Anymore
Intelligent Emergence: What's the scale of your entire Series B financing? Is this the largest round of financing since the establishment of your company?
Yang Longsheng: Yes. This is our Series B2 financing, worth 150 million yuan. Our entire Series B financing is divided into three parts, and the total financing amount may reach nearly 300 million yuan.
Intelligent Emergence: The glasses market in China has been fluctuating. How did this wave of AI glasses become popular?
Yang Longsheng: OpenAI and Deepseek have driven a wave of AI enthusiasm. People's passion for AI has awakened the demand for hardware products or application scenarios. Glasses have caught this wave of enthusiasm. They are the best carriers for AI, being closer to people and allowing for quick access to AI.
Intelligent Emergence: A few years ago, only state - owned enterprises were interested in AR glasses companies, but the industry has become much more active this year. As an industry insider, what's your specific feeling?
Yang Longsheng: Yes, last year, only state - owned enterprises were investing in the industry. This year, some strategic investors from various industries and market - oriented funds have also gradually increased their investment in this field. Through the introduction of some friends and old shareholders this year, there should be over a hundred investors who have approached us. Our Series B2 financing really started in April or May, and the whole transaction period was completed in just over two months. It was quite fast.
Intelligent Emergence: How popular do you think AI glasses are? Are there any impressive details?
Yang Longsheng: The simplest sign is that many people around me who didn't pay attention to the smart glasses industry before have come to ask me, "Is it worth buying Xiaomi's glasses?" Several leading internet companies may launch their products at the end of the year.
Intelligent Emergence: What are the core questions of these investors? Are there any common points?
Yang Longsheng: The core questions are all about what we think of the AI glasses market, what impact Xiaomi's entry will have on us, and what our competition strategies are against tech giants in the future.
There Is Still a Window Period of Two to Three Years for Startups After the Entry of Tech Giants
Intelligent Emergence: I think the "Four AR Dragons" seem to be more committed to their main positions and haven't removed the "AR" label like other manufacturers. How do you think about this? Have you ever wavered?
Yang Longsheng: We always believe that AI glasses will gradually evolve towards the direction of having a display function. Looking at the overall market, in the past, the market share of smart glasses without a display function and mainly focusing on audio was very low, which proves that for consumers, glasses with a camera and a display function are more in line with the public's psychological expectations for AI glasses.
Intelligent Emergence: After the entry of tech giants, how do you think startups in this field should deal with the competition?
Yang Longsheng: I think tech giants are currently in the trial - and - error stage, and there is still a window period of two to three years left for startups to become leading brands. Currently, we are ahead of tech giants in terms of both market share and product leadership.
Now this market is becoming more concentrated at the top. Investors still prefer to invest in the first - and second - ranked companies in the industry, showing a trend of polarization.
Intelligent Emergence: What's the user profile of your products?
Yang Longsheng: We focus more on the new middle - class and new - consumption groups. These people are mainly in first - and second - tier cities and pursue personalized brands, which allows us to compete with tech giants like Xiaomi and Huawei at different times.
Intelligent Emergence: The competition in the AI glasses field may be very fierce in the future. What's your core competitiveness?
Yang Longsheng: There are several key points. Firstly, in terms of optics, we are the first company to achieve mass production of waveguide products, and we have also achieved binocular full - color and a large FOV (Field of View).
At the OS operating system level, we have integrated AI into the middle layer of the underlying system to support existing applications and achieve AI - native features. For example, you can use the glasses to order food or buy things automatically. We also focus on exploring application scenarios in social and entertainment fields.
Finally, in terms of the supply chain, we have been insisting on localization. Now the localization rate has actually reached over 80%, which gives us a cost advantage. Our glasses can now be produced at a cost similar to that of photo - taking glasses. In terms of channels, we have also opened the first offline brand store with Asia Optical.
Intelligent Emergence: How much has the overall cost of smart glasses decreased?
Yang Longsheng: Compared with before, it may have decreased by about 30%. We have been working on the design of the array waveguide optical path and the optimization of material processes, and the yield rate has increased by nearly 15% compared with before.
Intelligent Emergence: What's the approximate shipment volume of the AI glasses market now? How about your sales?
Yang Longsheng: This year is better than previous years. I estimate that the volume of the entire smart glasses market in the Chinese market should reach six to seven hundred thousand units this year, which is about two and a half to three times more than last year.
The global market is relatively large. There may be nearly millions of units shipped this year, with most of the volume being taken by Meta. We predict that Xiaomi's shipment volume this year should reach 300,000 to 400,000 units. We should also rank second this year, with a sales volume of about 150,000 units.
Although the Threshold of AI Glasses Is Low, There Are Many "Pitfalls"
Intelligent Emergence: I feel that the threshold of the AI glasses field is very low. For example, many AI glasses solution providers have emerged in Huaqiangbei.
Yang Longsheng: In terms of the underlying logic, it's not that difficult to make a camera in the form of glasses. However, it's very difficult to integrate AI functions and provide a good overall interaction experience.
It actually took Xiaomi nearly a year to develop its AI glasses, and there are too many details to optimize.
For example, if you want to create an AI + translation scenario, you need to first optimize the latency. We have developed some algorithms to ensure that the glasses can accurately recognize what I'm saying and translate it for the other party. Even for a simple phone - call problem, I need to solve the problem of directional sound transmission of the microphone.
Another example is that if we want the glasses to be wearable for a whole day, we need to use a dual - chip solution - using low - power Bluetooth for daily processing and a high - performance SoC to solve the problems of photo - taking and input. The dual - chip solution also involves many problems with Bluetooth communication protocols. There are many pitfalls, all of which are technical difficulties in software and hardware R & D. I think it's a comprehensive competitiveness.
In the future, AI glasses need to continuously improve their 6DoF capabilities, full - color display, and lightweight design, which require high - end hardware experiences. I don't think manufacturers in Huaqiangbei can participate in this, as it requires a huge investment. The products from Huaqiangbei are just security cameras in a glasses shell.
Intelligent Emergence: In the short term, AI glasses seem to be quite homogeneous. What's the root cause of this?
Yang Longsheng: The imagination space of AI is relatively large, and the application scenarios are emerging with the development of technology. However, currently, the main scenarios being solved are concentrated in meetings, translation