Exclusive Interview with Thunderbird Innovation CEO Li Hongwei: Telecom Giants Bet on the Next Generation of "Mobile Phones"

Author | Huang Yu Editor | Zhou Zhiyu 2025 will be a breakout year for AI applications, and smart glasses—seen as one of the best carriers for large AI models—are undoubtedly among the hottest tracks in the tech industry. From Meta driving rapid sales growth of Ray-Ban Meta, to the sudden heat of the "Hundred Glasses Battle" in China, from startups to major internet companies and smartphone manufacturers, virtually everyone is talking about AI/AR glasses. The rapid influx of capital has quickly brought this once-quiet industry into the spotlight. However, behind the "Hundred Glasses Battle," very few products actually achieve sustained shipments and user retention. Shipment volume, channel capabilities, and technological accumulation are becoming the dividing line, with capital and resources increasingly concentrating among a few leading companies. At this moment, on the second working day of 2026, leading AR glasses manufacturer Thunderbird Innovation announced a new round of financing, jointly invested by funds under China Mobile and China Unicom. This also marks the first time that China’s mainstream telecom operators have jointly invested in the smart glasses sector. "In the past year, the number of investors entering the smart glasses sector has clearly increased," revealed Li Hongwei, founder and CEO of Thunderbird Innovation, in an exclusive interview with Wallstreetcn. Thunderbird Innovation raised far more than expected this round—two to three times the original plan. At the same time, he pointed out, although investor enthusiasm is high, funds are still gathering at the top. Counterpoint data shows that Thunderbird Innovation held a 24% global market share in Q3 2025, ranking first in the global AR smart glasses market for two consecutive quarters. Additionally, according to Wallstreetcn, last year Thunderbird Innovation's overseas revenue share exceeded 30%, and is expected to surpass 50% in 2026. Globalization is a key strategy for the company, with North America, Europe, Japan and South Korea, Australia, the Middle East and Southeast Asia all part of their current and future primary market layout. Regarding the reasons why the major Chinese operators have chosen to jointly bet on smart glasses at this time, Li Hongwei believes that the industry’s development trend is now very clear, and everyone sees the sector is growing rapidly and products are maturing quickly. "Operators judge that the industry is at a tipping point for explosive growth." From the operators' own development perspective, their investment in Thunderbird Innovation is based on two main considerations: the first is strategic—the operators want to focus on the track of the "next-generation mobile phone"; the second is their belief that now is a good time to cooperate with smart glasses manufacturers. Notably, with cash investments from China Mobile and China Unicom's funds, Thunderbird Innovation will achieve important breakthroughs in products and channels—during CES 2026, Thunderbird will unveil its first eSIM AR glasses, featuring full-color binocular display and built-in eSIM module. "Since smart glasses are the next-generation mobile phone, they need to be relatively independent," Li Hongwei noted. “Whether it’s color MicroLED or eSIM, only Thunderbird is truly building the 'next-generation mobile phone.'” Thunderbird’s partnership with China Mobile and China Unicom is systematic; on one hand, they will co-develop products (besides eSIM, operators could also help Thunderbird lay out edge computing, offsetting shortages in computing power and latency for smart glasses); on the other, operators' massive user base and channel capacity provide a realistic path for scaling Thunderbird's products. "More than a decade ago, operators drove the rapid adoption of smartphones—top-ups came with free phones and the spread was very fast," said Li Hongwei. Thunderbird plans to partner with operators to launch top-up promotions with free glasses. Once mature, this model may become an important inflection point for industry penetration. There’s no doubt capital investing in smart glasses represents a contest for the "next-generation mobile terminal" and "AI super entry point." The potential for smart glasses’ future development is now widely recognized; the debate is simply about timing—and who will bring smart glasses to millions of homes. "I don’t really agree with the idea that 2025 is the explosive or breakthrough year for smart glasses," said Li Hongwei. Over the past year, there’s been no fundamental breakthrough in product power; the industry is lively, attracting more people to the market, but attention to experience standards is still lacking. Li Hongwei asserts that 2026 will be the true 'year one' for smart glasses, and 2027–2028 will be the 'iPhone moment.' There is a difference—2026 will be when products meet user experience standards, and 2027–2028 achieve the state of the first-generation iPhone. This assessment contrasts with much of the optimism in the market. Over the past year, smart glasses frequently took the spotlight, some functions went viral, boosting attention and prompting many to call 2025 the "year one" or "breakthrough year" for smart glasses. Clearly, for Li Hongwei, excitement doesn’t equal maturity. Many smart glasses products resemble mere combinations of features; there remains a gap before ordinary users are willing to wear them all day, long-term. This also explains why many products gain short-term attention but struggle to build long-term competitiveness. Li Hongwei divides the coming industry evolution into three stages: The first stage runs from now to before the "iPhone moment," with many players—the core is who can truly create the next-generation "iPhone." The second stage goes from the iPhone to the iPhone 4, as products go from revolutionary definition to mass adoption. The third stage, from the iPhone 4 era onward, is when surviving players engage in full competition across all dimensions. On the eve of industry eruption, all players must go all out to become the first company to craft the "iPhone" of smart glasses. Li Hongwei points out, at this stage, the fundamental logic is who is determined and has enough resources—money, people—to run fast enough. According to Wallstreetcn, Thunderbird Innovation’s latest cash financing round has surpassed 1 billion yuan. Li Hongwei stated that the core use of these funds is R&D—to enable products in 2026 to quickly reach experience standards and expand competitive advantages in optical display, spatial computing, AI and other key technologies. "Last year, many products were launched in the industry, but frankly, almost none truly met user experience standards," he said. Whether users feel good about it after use, and would recommend it to others, is his core standard for product maturity. Nevertheless, Li Hongwei believes that whether for new technologies or past engineering investments, now is the time when sustained effort on these two fronts will meet experience standards. Returning to the product itself, currently external definitions of AR glasses and AI glasses are not uniform; their technical routes, functionality, and market promotion often overlap or blur. Some vendors position AI glasses as a transitional form of AR glasses; some often define their products as "AR+AI" glasses. Li Hongwei believes future smart glasses will mainly take the following forms: First, AI glasses without displays—this track is closer to the logic of some major companies’ past products, and the main strategic value is data collection to enhance the existing mobile AI ecosystem. Second, AR glasses with displays, subdivided into two categories—one resembles regular glasses with minimal features, and the other is more powerful but heavier and costlier. Technologically, Thunderbird chooses the relatively "heavy" path: full-color MicroLED waveguide display, low power consumption architecture, deep AI and spatial computing integration. This is a difficult but correct route, seen by Li Hongwei as the necessary path toward the "next-generation terminal." On the first working day of 2025, Thunderbird Innovation shocked the industry by announcing a strategic partnership with Alibaba Cloud—the Tongyi large model series will provide exclusive, customized AI technology and product support for Thunderbird's AI glasses. At the time, Li Hongwei revealed that exploring the Killer APP would be the top priority in their cooperation, while on-device large models ranked second or third. Now, Li Hongwei says that building the killer app remains Thunderbird’s most important pursuit, and they are nearly ready for launch. On this steady, snow-covered slope of the smart glasses track, excitement is just the beginning. The real watershed likely still lies ahead. **Below is the full edited dialogue between Wallstreetcn and Li Hongwei:** ## Smart glasses industry reaches explosive inflection point **Wallstreetcn: Why do you think operators chose to invest in Thunderbird at this point? How do they see the industry’s outlook?** Li Hongwei: First, it's a two-way process, because everyone shares the same conclusion about the industry. From the operators’ perspective as investors, they see the industry at a tipping point for explosive growth. Though the "iPhone moment" hasn’t arrived—I think that should be 2027 or 2028—the direction is clear and they see rapid growth and product maturity. For operators, their considerations are twofold. One is strategic: what is the next-generation phone, and how they lay out for the future. That’s a capital perspective. Second is business cooperation—when will products be mature enough for such partnerships? Operators face hundreds of millions, even billions of users, and want to provide good product experiences. These two judgments both point to now being a good time. **Wallstreetcn: Many say this year is the explosive or breakthrough year for smart glasses. Do you agree? What new trends have emerged over the past year?** Li Hongwei: Frankly, I don't quite agree. I don't think 2025 is the year one or breakthrough year for AI glasses. When will it be? Possibly 2026, because then products will actually exceed user experience standards. The change in 2025 is that more people are getting into the field, but what matters is not just numbers but actual breakthroughs—the underlying logic. Many jump in after seeing market opportunity, but not the market’s substance. Even when discussing the "year one," people don’t see the essence. If I must choose, I think 2026 is year one; 2027–2028 the iPhone moment—there is a distinction. **Wallstreetcn: You mentioned "many people not seeing the essence of the market"—what do you mean?** Li Hongwei: People see the surface excitement, not the underlying change. Few discuss "experience standards," even inside the industry. Many products are just fun combinations, but the real question is whether you meet experience standards. Some become pessimistic, say it’s not worth doing, but I think they’re missing the point. They haven’t seen that experience standards will fundamentally change in 2026. **Wallstreetcn: How do you judge the future evolution of the competitive landscape?** Li Hongwei: I divide the industry into three phases: Phase 1: From now until the iPhone moment, you’ll see all sorts of players—key is whether you can really build an "iPhone." Phase 2: From "iPhone moment" to "iPhone 4," major firms formally enter, product leadership remains critical. "Formally entering" means not just dabbling: for example, is Xiaomi really in now? I don't think so, judging by their product results. They haven't put in their main team or core resources; if they had, results should be better. Phase 3: The remaining players battle it out across all dimensions. In phase one, the key is who has the resolve and resources—money, talent—to run fast and be first. ## Why did operators choose Thunderbird? **Wallstreetcn: What do operators value most in Thunderbird’s core competitiveness?** Li Hongwei: From a strategic viewpoint, it’s about who can ultimately succeed in this arena. They judge that Thunderbird is most likely; from a product perspective, they believe Thunderbird offers the best products on the market. Thunderbird’s products have several strengths: First, a large user base—recent Counterpoint Q3 numbers show we’re number one globally in AR glasses, with a 24% share. One in four AR glasses sold globally is from Thunderbird. Large user numbers mean the product is proven. Second, our product experience standards—user feedback is pretty good. Third, our products are priced appropriately for the category. Operators previously offered phones with top-ups; now glasses require the right price range. Top-up promotions with free glasses mark the start of an inflection point. As for why they judge Thunderbird as most likely to succeed: Compared to others, our key advantage is leading product technology. For any new category, you must deliver valuable products—backed by technology and product leadership. They see this, especially our leadership on the right path. Globally, only Meta and Thunderbird are truly building the next-generation phone—a full-color AR glasses all-in-one. Also, if they invest, they want to invest in startups; among them, Thunderbird leads in the Matthew effect. From a financing perspective, it’s clear that top, promising companies attract the bulk—operators see this too. Ultimately, it’s also about team and strategy. **Wallstreetcn: Besides funding, what other support or cooperation will China Mobile and China Unicom bring to Thunderbird?** Li Hongwei: There will be systematic cooperation, mainly in two areas— First, co-development: for example, adding eSIM to Thunderbird glasses. Previously, glasses had to be paired with a phone via Bluetooth, which isn’t ideal. That approach depends on the phone, but glasses should be independent, so only Thunderbird is really making the next-gen phone, whether it’s color MicroLED or eSIM. Besides eSIM, there will be broader cooperation—AI is increasingly important. Through operator partnership, we don’t just deploy in the cloud but also at the edge. Operators have extensive AI compute resources at base stations, so our AI can run faster—doing everything in the cloud leads to network latency. For example, users can upload photos and videos to cloud albums, enabling many business collaborations. Second, user promotion: Over a decade ago, operators drove the rapid spread of smartphones by offering phones with top-ups. Promoting the glasses using the same model covers a huge user base—and also leverages operators’ brand trust. **Wallstreetcn: When will you launch the top-up-with-glasses promo? What are the specific plans?** Li Hongwei: We know approximately which price segments; timing is still undecided, so it’s not ready for public disclosure. ## Plans for independent IPO in the future **Wallstreetcn: This financing reached 1 billion yuan. What will you focus investments on?** Li Hongwei: The main core is R&D. The problem facing industry products is that almost none meet user experience standards. After using the product, do users feel it’s great, willing to recommend it, with no complaints, high maturity? Not yet. Thunderbird has built leading technology over the past four years, so we’re clear about current tech standards. For both new technology and past engineering investment, now we’ve reached the point where sustained effort in these two areas will meet experience standards. Our funds are truly for substantial R&D to help 2026 products quickly meet those standards. **Wallstreetcn: How has the financing environment in the sector changed over the past year? What’s next?** Li Hongwei: First, there are clearly more investors—so we oversubscribed a lot. The funds raised were 2–3 times our original goal; we even turned down offers of similar size. Investor enthusiasm is much higher. Second, funding is converging toward the leading companies—investor interest is high but it’s the top companies that actually raise significant funds. The concentration is clear, as the industry is at this stage. Before the iPhone moment, funding first aggregates to the top, and afterward may diffuse a bit, but the overall trend is concentration. **Wallstreetcn: Why does investment scatter after the iPhone moment?** Li Hongwei: Because after the iPhone moment, confidence grows and you can’t invest in the very top, so you may look for others. At the current stage, you can still invest in the top startups. **Wallstreetcn: Any plans for independent IPO in the future?** Li Hongwei: Yes, that's our plan, but there’s no public timetable yet. ## Overseas revenue proportion to surpass domestic this year **Wallstreetcn: What is Thunderbird’s overseas business proportion now?** Li Hongwei: Our overseas share is now above 30%; in 2026, overseas will basically surpass 50%. Thunderbird is inherently a global company; doing smart glasses means going global. Today’s entrepreneurs see the world as one market—we’re much more global in mindset. Thunderbird lays out strategic focus market by market, not spreading resources thinly. We started with China, then entered the US, focusing on the US from mid-2024, so during Black Friday and Christmas you’ll see our expanded presence in North America, Europe, Japan, and Korea. Thunderbird is among the best-selling products on US and European Amazon, which is part of our layout. Through these channel arrangements, in 2026 we’ll have solid layouts in North America, Europe, Japan, Korea and other countries, so overseas revenue will naturally exceed domestic. **Wallstreetcn: What's the core strategy for market expansion?** Li Hongwei: If we can win in China and the US, we can replicate success elsewhere. If you can lead, just enter more markets. Of course, products need some customization, as needs vary by country. We adjust glasses accordingly, with differentiated software, marketing and branding. When going global, we didn’t follow the old path of overseas expansion—we quickly moved to localizing, with tailored marketing and branding strategies. For example, we focus more on “seeding” in China, branding in overseas markets—we plan to launch IP co-branded products in North America in early 2026. **Wallstreetcn: Are channels different domestically and abroad?** Li Hongwei: Channels differ. We follow two main rules—the most important is local channel layout. China is mainly online now for consumer electronics, but the US is still mainly offline, about 70/30 or 60/40. Though we started online in the US, we’re quickly entering offline—Walmart, Best Buy, Costco, Target, Sam's Club, etc. That’s the main difference. **Wallstreetcn: Are offline stores mainly big-box retailers or self-run flagship stores?** Li Hongwei: We’ll start with big-box stores, leveraging existing systems. We will have flagship stores later, but short-term their main use is brand building and user experience. ## Making killer apps the top priority **Wallstreetcn: Which technical directions is Thunderbird focusing on conquering?** Li Hongwei: We’re focusing on several technical directions: First is optical display—so next year’s glasses will feature displays. The main optical problem is making display imperceptible: a MicroLED waveguide glasses currently have issues, e.g. single-color glasses make the world look yellowish, with low light transmission. For those without myopia, it feels like vision is blocked, and rainbow artifact is an issue. From the outside, glasses seem to leak light. Second, performance issues—reducing power consumption for all-day wear, alongside optimizing brightness, resolution, etc. Second is AI: AI is seen as one of the key application scenarios for glasses, but most devices just transplant AI onto glasses, maybe with some speed optimizations; still, no breakthrough. We hope to build killer apps for AI on glasses—current apps just answer questions as on phones. With large-model training for long-term memory and core scenarios, we aim to turn them into killer apps. Third is spatial computing algorithms. Previously, little use was made of spatial computing—for example, live translation where the result is overlaid on signs as you walk; we're developing that. We not only build the algorithms but also link them to chips and device optimization. Fourth, basic experience improvements—hardware, software, battery energy density, physical size, weight, etc. From a tech perspective, those are the main directions. For users, it’s about comfortable wear and killer AI apps with high experience standards. **Wallstreetcn: Last year you said building killer apps was the top priority. Has that changed?** Li Hongwei: No, it’s still the most important thing. **Wallstreetcn: After a year, how far is your killer app development—what stage is it at?** Li Hongwei: We’re basically ready and close to launch. We've done a lot of fundamental work. **Wallstreetcn: Recently, Doubao phone spurred a heated discussion about AI Agent and data privacy. Some believe smart glasses may become the main terminal to realize true AI Agent—what's your view?** Li Hongwei: I agree. Smart glasses are native AI devices, without the legacy burdens of the mobile ecosystem, and can create value during non-screen time. Historically, successful disruptions come from the side, not head-on—smart glasses are AI’s best carrier: they see what you see and hear what you hear, and are native AI devices. There’s no need to battle entrenched interests in legacy ecosystems. **Wallstreetcn: When it comes to on-device AI models, does the ability differ between glasses and phones? How long will the gap persist?** Li Hongwei: Yes. For glasses, it’s not only compute but power consumption. I think closing the gap will take some time, since glasses must be comfortable and can’t have large batteries. So AI on glasses is best served by device/edge/cloud collaboration. Some processing can be done on-device: currently, glasses’ chips can run a 1b model, phones can do 6–7b. Power consumption is the key issue. **Wallstreetcn: Can mobile apps be authorized for use on glasses? Can your Agent directly operate commands in glasses?** Li Hongwei: The Agent ecosystem for glasses isn’t the same as for phones; glasses begin with a distinct ecosystem. We look at our phones for 6–7 hours daily, sleep 8, and have another 8 hours when we’re not looking at screens—AI can add value during those hours, making glasses the ideal entry point. So the ecosystem is different. When glasses need to leverage the mobile ecosystem, so far there’s been no conflict, like apps blocking Doubao phone’s assistant. Because glasses haven’t yet penetrated the phone’s core territory, the mindset is more open. Many have opened Agent API interfaces—so collaboration is easier. **Wallstreetcn: What stage is the smart glasses application ecosystem at now?** Li Hongwei: Progress is slow. True ecosystem building is not about building for the sake of it, but making great products. If your product is good, software makers will want to join—you make an iPhone, and even the first version without an App Store attracts developers. Building a broad ecosystem now is pointless—glasses lack enough users for tens of thousands of apps. **Wallstreetcn: When smart glasses have their iPhone moment, will software makers immediately join in?** Li Hongwei: Yes; right now, it’s about finding a killer app—you need some ecosystem interaction for that. Building a broad ecosystem now is useless—without users, too many apps add no value. **Wallstreetcn: How far along are you in enabling AI Agent capability on glasses?** Li Hongwei: We already have efforts underway, but not released yet. The Agent currently uses API interface methods, not GUI-Agent methods (GUI-Agent is less efficient, a last resort). We have GUI options, but prioritize API. **Wallstreetcn: When will your killer app have an effect similar to Doubao phone, impressing everyone?** Li Hongwei: We expect this effect with our next product launch—should be released in the first half of 2026. Earlier launches didn’t deliver on “excitement,” but we’re confident this time. --- Risk warning and disclaimer The market carries risks; investment should be prudent. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation or needs of any individual user. Users should consider whether any opinions, views or conclusions in this article are suitable for their particular circumstances. Investments made based on this article are at your own risk.