Hello, ladies and gentlemen. Thank you for standing by. Welcome to Hesai Group's Third Quarter 2025 Earnings Conference Call. Please note that today's conference call is being recorded. I will now turn the call over to our first speaker today, Yuanting Shi, the company's Head of Capital Markets. Please go ahead.

Thank you, operator. Hello, everyone. Thank you for joining Hesai Group's Third Quarter 2025 Earnings Conference Call. Our earnings release is now available on our IR website at investor.hesaitech.com as well as via Newswire services. Today, you will hear from our CEO, Dr. David Li, who will provide an overview of our recent updates. Next, our CFO, Mr. Andrew Fan, will address our financial results before we open the call for questions. Before we continue, I refer you to the safe harbor statement in our earnings press release, which applies to this call as we will make forward-looking statements. Please also note that the company will discuss non-GAAP measures today, which are more thoroughly explained and reconciled to the most comparable measures reported under GAAP in our earnings release and filings with the U.S. Securities and Exchange Commission and the Hong Kong Stock Exchange. With that, I'm pleased to turn over the call to our CEO, Dr. David Li. David, please go ahead.

Thank you, Yuanting, and thank you, everyone, for joining our call today. Let's start with an overview of this quarter's progress. Q3 was a quarter of powerful momentum and exceptional execution with net revenue surging nearly 50% year-over-year and a landmark milestone achieved, we produced over 1 million LiDAR units in 2025 alone and are the first to do so globally. We've also led the long-range automotive LiDAR market for 7 consecutive months, capturing an impressive 46% share in August, 1.5x the second player and 2.4x the third according to Gasgoo. Our profitability performance is even more remarkable. After turning solidly profitable ahead of schedule in Q2, we kept the momentum going in Q3, delivering a record quarterly GAAP net income of RMB 256 million and a 9-month GAAP net income of RMB 283 million, achieving our full year target of RMB 200 million to RMB 350 million, well ahead of schedule. This milestone further reinforces our undisputed financial leadership in the LiDAR industry. With robust growth and solid profitability working hand-in-hand, we're building powerful long-term momentum and creating sustainable value for our shareholders. Now let's dive into our Q3 business highlights. Starting with our progress in the ADAS market. Firstly, for ADAS, LiDAR is no longer optional. It's rapidly becoming a standard feature. As a result of our product leadership and strong client relationships, we are proud to announce our new design wins from both of our top two ADAS customers across all their 2026 models, achieving 100% LiDAR adoption. On top of that, select facelifted versions of Zeekr's flagship models are now rolling out with Hesai LiDAR as a standard feature. Looking ahead, a growing number of best-selling models across our diverse client base are slated for SOP with Hesai in the second half of 2025 and throughout 2026, further cementing our position as a LiDAR partner of choice. Beyond this, we are excited to see China taking decisive steps towards higher-level autonomous driving. In September, the MIIT introduced conditional approval for L3 vehicle production for the first time. This was quickly followed by a public consultation on the new mandatory safety standard for L2 systems. Together, these regulatory developments are clearing the runway for a new era of smarter, safer autonomous driving in China. As regulations take shape, one thing is clear, a higher-level autonomous driving system cannot tolerate a single point of failure, making safety redundancy not just important, but essential. At the same time, LiDAR sensors must be factory integrated rather than retrofitted, pushing automakers to future-proof their platforms for tomorrow's L2 and L3 capabilities. The trend is accelerating even as software capabilities continue to evolve, pioneering OEMs are already launching multi-LiDAR vehicles in 2025. These models featuring 2 to 5 LiDARs are winning consumer recognition and achieving strong sales results. To gear up for the new era of L3 autonomous driving, we launched our Infinity Eye B LiDAR solution in April. It pairs our forward-facing long-range ETX LiDAR, a new benchmark with the world's longest detection range, with FTX blind-spot LiDARs, offering the industry's widest field of view. Most excitingly, I'm thrilled to share that this quarter, ETX landed yet another design win, this time with a top three domestic new energy vehicle automaker, one of our valued existing customers, paired with multiple FTX units for full 360-degree blind-spot coverage. Mass production is slated for late 2026 or early 2027. These developments reaffirm a principle we've always stood by. The cost of LiDAR is nothing compared with the priceless value of human life. As the auto industry moves toward higher-level autonomy, LiDAR content in new vehicles is ramping up fast. We now expect 3 to 6 LiDARs per L3 vehicle, representing a system value of roughly USD 500 to USD 1,000 per car in the long run. This trend is massively expanding our addressable market and supercharging the long-term growth potential of our ADAS business. Beyond our progress in ADAS, our Robotics business is becoming an increasingly powerful growth driver, fueled by expanding autonomous driving fleets. As core technologies advance rapidly, autonomous driving companies worldwide are approaching a tipping point towards scaled operations, and we're proud to be a key enabler of this transformation. In China, the latest generation of autonomous driving fleets are adopting ADAS LiDAR solutions alongside optimized chips and vehicle design to lower total BOM costs and accelerate commercialization. Spearheading this shift, we've recently signed new deals with Pony.ai, Hello Inc. and JD Logistics. And I'm excited to share that for some of their models, all LiDAR units, up to 8 main and blind-spot LiDARs will be supplied entirely by Hesai. Internationally, we have also made strong progress. Many overseas robotaxi companies continue to favor mechanical spinning LiDARs for their performance and stability, making them less price sensitive and creating meaningful opportunities for us. We are proud to share that we have signed new LiDAR supply agreements with leading global autonomous driving companies, including Motional and others across North America, Asia and Europe. These large-scale programs represent deals worth tens of millions of dollars, with strong follow-on potential as deployments expand. As our partners move toward large-scale operations in the coming years, this marks a defining milestone for the autonomous driving industry. Building on these operational milestones, September marked a historic moment for Hesai as a public company. We successfully listed on the Main Board of the Hong Kong Stock Exchange, becoming the world's first LiDAR company with dual primary listings in both the U.S. and Hong Kong. This was the largest IPO in the global LiDAR sector, raising USD 614 million after the greenshoe option, with strong support from global institutional investors and industry leaders. The offering underscores confidence in the long-term potential of the LiDAR industry and in Hesai's ability to deliver at scale. More importantly, it strengthens our financial foundation, enabling us to invest in innovation and capture new market opportunities. To wrap up, our strong Q3 results are a powerful testament to Hesai's momentum and execution. The successful completion of our Hong Kong IPO marks a bold new chapter for Hesai. We are witnessing the dawn of an AI-driven fourth industrial revolution, an era that promises unprecedented gains in productivity and human well-being. As we look toward the decade ahead, Hesai is rising to this moment, evolving into a full spectrum technology infrastructure builder that redefines how cars and robots perceive and interact with the world. With that, I will now turn the call over to Andrew to share more details on our financial performance and outlook. Andrew, please go ahead.

Thank you, and hello, everyone. Before we get into our financial performance this quarter, I'd like to start with a key milestone for Hesai as a public company. In September, we completed our dual-primary listing on the Main Board of the Hong Kong Stock Exchange under the ticker, 2525. Through this global offering, Hesai has become the world's first LiDAR company to be listed in both the U.S. and Hong Kong capital markets. The market response to our Hong Kong debut was exceptional. The public tranche was nearly 169x oversubscribed, while the international tranche attracted demand of more than 14x the available shares. In total, we raised USD 640 million after the greenshoe option, further strengthening our balance sheet and improving trading liquidity. These resources have greatly enhanced our capacity to invest in innovation, expand production and drive operational excellence. We are now in a strong position to capture growing opportunities in the global LiDAR market and extend our leadership as adoption continues to accelerate. I will now walk through our Q3 financial and operational performance. To be mindful of the length of our earnings call today, I encourage listeners to refer to our earnings release for further details. Q3 was another outstanding quarter, delivering record-breaking results across the board. Total shipments reached 441,398 units, up 229% year-over-year, while net revenue surged 47% to RMB 795 million or USD 112 million, marking our sixth consecutive quarter of robust year-over-year growth. This powerful momentum fueled by the surging adoption of our category-defining ATX amid the industry's rapid shift toward LiDAR as a standard feature along with a 14-fold year-over-year rise in robotics LiDAR shipments across expanding applications underscores the strength and scalability of our business model. Our gross margin remained healthy at 42%, driven by economies of scale and continued gains in manufacturing productivity. Just as importantly, we're now embedding AI across R&D, operations and customer support, unlocking new efficiencies and strengthening the foundation of a lean optimized expense structure. You may have noticed that today's prepared remarks are being delivered through AI-generated voices. While the pronunciation isn't perfect yet, it's a small but meaningful example of our commitment to wholeheartedly embrace AI across the organization. We believe that for companies today, embracing AI is just like embracing digital transformation 20 years ago. It's the key to building greater competitiveness for the future. Since Q2, we've deployed an intelligent assistant across a wide range of daily workflows, cutting costs, shortening cycles and improving quality. This AI-driven approach has already delivered tens of millions of RMB in savings across travel, documentation, hiring, testing, coding and more. As a result of our adoption of AI and other cost control measures, total operating expenses declined year-over-year in Q3, keeping us on track to achieve RMB 100 million in OpEx savings in 2025 compared with last year. Building on our strong momentum, we delivered a record net income of RMB 256 million or USD 36 million in Q3, bringing our 9 months total to RMB 283 million or USD 40 million. We've already hit our full year profit target of RMB 200 million to RMB 350 million, one quarter ahead of schedule. This achievement reflects the scale and efficiency our business has reached, where growth is now translating directly into earnings. Higher volumes drive better unit economics, which in turn fuels more growth, creating a self-reinforcing cycle of profitability and innovation. It's worth noting that Q3 net income included gains from equity investments of RMB 148 million or USD 21 million. Excluding these gains, quarterly net income would have remained strong at RMB 108 million or USD 15 million. Taking this into account, we are raising our full year GAAP net income guidance for 2025 to a range of RMB 350 million or USD 49 million to RMB 450 million or USD 63 million, and we expect full year net income, excluding these gains from equity investments to stay within our earlier guidance range of RMB 200 million to RMB 350 million. For the remainder of the year, we expect to carry forward the strong momentum we have built. For Q4, we're projecting net revenues of between RMB 1,000 million or USD 140 million and RMB 1,200 million or USD 169 million, representing a year-over-year increase of 39% to 67%. To wrap up, our successful listing on the Hong Kong Stock Exchange marks an exciting new beginning for Hesai. We're growing faster, scaling smarter, and executing stronger than ever with accelerating revenues, solid margins and proven profitability. We're building competitive advantages that will keep compounding over time. We are more energized than ever to seize the opportunities ahead. This concludes our prepared remarks today. Operator, we are now ready to take questions.

Your first question comes from Tina Hou from Goldman Sachs.

Congratulations on a very strong result. So my question, the first one would be related to the pricing side of things. As we go into the last quarter of this year and enter into price discussion with customers next year. So just wondering if you could give us any color in terms of what kind of pricing we're looking at for next year, considering everything, the annual price cut, competitor dynamics. That's number one. Number two is in terms of the volume. So just wondering if management is seeing next year, the OEM customers are going to accelerate the adoption of LiDAR. And in your view, when will be sort of the starting point or the takeoff point for mass market models to start having LiDAR as a standard option? And I guess related to that, if there is any color you can give us in terms of your 4Q as well as 2026 guidance or any kind of color on volume, it would be really helpful.

Thank you, Tina. So I understand your question. I will try to cover our guidance or forecast or color for the current year and also year 2026. Let's talk about the 2025 year, full year guidance first. On revenue side, our Q4 revenues are expected to reach about RMB 1 billion to RMB 1.2 billion, bringing full year 2025 revenues to approximately RMB 3 billion to RMB 3.2 billion, representing a year-over-year increase of nearly 50%. This strong growth is driven by the rapid adoption of LiDAR in passenger vehicles and the expanding use of the robotic LiDAR across new applications. Volume and ASP. During the first three quarters of 2025, we shipped about close to 1 million units in total. We expect the shipments to continue accelerating throughout the year with Q4 shipments reaching approximately 600,000 units as a seasonal high. The ATX LiDAR is expected to account for roughly 80% of total deliveries in Q4 in terms of volumes. It has a market price of around $200 with discounts offered to major customers on our pricing strategies. The stronger-than-expected demand for ATX has accelerated its replacement of the AT128 LiDAR among our OEM customers in the second half of 2025. Meanwhile, several automakers have adjusted their second half production schedules for vehicle models equipped with AT128, leading to softer demand for the product. As AT128 is priced at several times the price of ATX, this shift in product mix has resulted in a relatively lower blended ADAS ASP for year 2025. Margin-wise, the blended gross profit margin is expected to remain healthy at around 40% in Q4. We are raising our full year 2025 GAAP net income guidance to RMB 350 million to RMB 450 million. Excluding gains relevant to equity investments recorded in Q4, normalized full year GAAP net income remains within our previous guidance range of RMB 200 million to RMB 300 million. On the non-GAAP metrics, you should add an additional RMB 120 million for stock-based compensation. Looking ahead for year 2026, we see it as a true inflection point. On one hand, we anticipate strong demand for ADAS LiDAR in passenger vehicles with our LiDAR shipments expected to reach at least 2 million to 3 million units or potentially even higher if L3 adoption becomes an industry-wide trend. On the other side, we do anticipate a potential decrease in blended ASP. That's mainly due to three things: one, a shift in product mix towards our ADAS LiDARs, which have a relatively lower unit price, but we will see higher deliveries and revenue share; two, the modest volume-based pricing for our large order strategic customers; three, the standard annual decline for downstream customers. That being said, there is reason to be optimistic. We expect a strong positive catalyst to emerge in year 2026 and 2027. First, L3 vehicles deployment in China will drive multi-LiDAR setups, pushing LiDAR content per vehicle to USD 500 to USD 1,000. We've already landed a flagship L3 program with a renowned customer and more exciting deals are in the works. Second, our overseas ADAS business is expected to start contributing, marking the beginning of global ADAS LiDAR mass production. Third, our Robotics business continues to gain momentum across diverse applications and customers, and it typically carries a higher ASP and margin compared to ADAS. Fourth, we are also exploring new growth engines, and we'll share more updates as things progress. On the profitability front, we expect gross margins to remain relatively stable in 2026 compared with 2025, supported by continued cost optimization across product and ASIC design, supply chain and manufacturing. At the same time, growing adoption of multiple LiDAR in ADAS is expected to help offset pressure on blended ASPs. In short, we expect to enter 2026 with a clear path towards a double-digit year-over-year revenue growth, accelerated shipments, a stable margin profile and potential new growth engines. Altogether, this sets the stage for sustained growth in the years ahead. Tina, hopefully, this covers your questions about our guidance in the next.

And I just want to say that was actually, Andrew. It's not a robot, and I am not a robot.

Your next question comes from Tim Hsiao from Morgan Stanley.

This is Tim. Congratulations on the robust results and steady project wins. I've got two questions. The first one is about the competition because we noticed that the competitors in China launched new products to undercut Hesai's ATX product. So how should we think about the peers, the mainstream product like EMX versus the key volume driver of the Hesai, i.e., ATX? The second question is about the technology because we noticed lots of discussions over the past few months about the SPAD SoC system and chip lately. So how should we think about the advantage of SPAD-based digital LiDAR? So I also want to get some updates from the management. Yes, those are two questions from my side.

Thank you, Tim. I appreciate the opportunity to provide more clarity on our competition and perspective on upcoming technologies. The market is indeed highly competitive, and we face significant challenges from several players, including one notable competitor with excellent products and technology. However, I want to broaden the view a bit. Our strategy involves a carefully structured timeline for releasing each product generation. In the mechanical LiDAR era, our Pandar128 was a leader in the field. We don't need to revisit that phase. Then came the AT128, which, looking back, clearly defined the automotive LiDAR industry, as we shipped the highest volume at a comparatively higher price point than competitors. Most importantly, it earned a reputation for being the highest quality and performing product available, which has contributed to its success. Now, we have Gen 3, the ATX, which we can confidently view as another market triumph. We have secured more contracts than our competitors and are shipping significantly larger volumes. The consensus is that the ATX demonstrates superior performance and reliability, aligning with the highest quality standards. We do have our roadmap for the next generation, but as a company that enjoys leading market share and a premium brand, we don't rush our releases just to keep pace with competitors. Our strategy focuses on maintaining a consistent rhythm and ensuring that all the best features are integrated into our products based on our timelines. This approach explains why, when competitors launch products six months to a year after us, they may include some interesting features, but overall, we have consistently achieved superior performance, volume, and margins. Thus far, our products remain the best-performing and most well-rounded on the market, and we hope to continue this trend. While we cannot predict the future, our strong semiconductor technology, manufacturing capabilities, brand power, and trusted relationships with top OEMs in China position us well for continued success. We expect price declines with each generation, as is typical in the market, yet we are committed to innovation to sustain gross margins. Moving on to your second question regarding SPAD, or Single Photon Avalanche Diode. I want to provide a deeper understanding beyond just competitive advertising. We were the first to implement SPAD technology in automotive LiDAR, having introduced the fully solid-state automotive LiDAR for near-range blind detection, the FT120, a few years ago. While our initial volume was modest, it is a fully automotive-grade product currently used in vehicles worldwide. Additionally, we acquired a company in Switzerland specializing in SPAD technology, as we believe their innovations will enhance our offerings. Nonetheless, we aim to remain objective about SPAD's capabilities today. A significant challenge with off-the-shelf SPAD technology integration is noise, leading to a higher likelihood of false triggers, which poses a safety risk for LiDAR products. As the industry seeks solutions to this problem, it’s essential to understand that while long-range SPAD offers high sensitivity, it also comes with an increased false-trigger rate that can jeopardize safety. We are actively developing an in-house solution to address these challenges. Our approach is not simply to incorporate the latest components for the sake of it; our priority is to include mature safety components first, followed by the latest technological advancements. Reliability and safety must come first, as we wouldn't want to compromise these critical aspects for a LiDAR system. We often liken our technology to an invisible airbag; it's crucial not to trade off safety for new features when there are known issues like false triggering. We believe these challenges will be resolved over time, as SPAD technology offers many promising features and cost advantages that will ultimately be integrated into our solutions, and we are working diligently toward that goal. I just want to ensure that everyone is well-informed about the advantages and limitations of this intriguing technology that is currently under evaluation.

Your next question comes from Jeff Chung from Citi.

David, this is such a great result and congratulations with the excellent earnings. So I have a question on L3 for David. So any sign for the improvement of the product mix and LiDAR per car? And any view on the Level 3 legislation in both China and Europe? And I also got a question for Andrew. So it looks like the 4Q guidance is really optimistic in the sense that in the best case scenario, revenue should up 50% Q-on-Q and the core earnings should up 100% Q-on-Q. Could you share with me the views of why you're so confident on this?

Jeff, okay. Let me first cover these two questions. For the legislation in relation to L3, we are thrilled by China's decisive push towards higher-level autonomous driving. As regulations evolve, safety redundancy becomes non-negotiable and the LiDAR must be factory integrated rather than retrofitted, driving automakers to future-proof platforms for L3 capabilities. The market is moving really fast. Leading OEMs are already rolling out multi-LiDAR vehicles in year 2025. Huawei's AITO M9 with four LiDARs, AVATR 12 with four, Zeekr 9X with five and NIO ES8 with three, all received strong customer demand and proving that demand for smarter, safer vehicles is real. So we see tremendous upside in LiDAR content growth. As L3 adoption accelerates, the number of LiDAR units per vehicle is expected to increase to three to six or even more, along with the trend towards upgrading main LiDARs to high-end models like our ETX. This could lift the total LiDAR content per vehicle to around USD 500 to USD 1,000. Beyond the numbers, every additional LiDAR unit directly enhances safety, underscoring the irreplaceable value of our technology and the critical role we play in shaping the autonomous driving future. We are seeing customers increased discussion for L3 applications and, of course, more LiDARs. We have signed a flagship program featuring ETX and multiple FTX already and more contracts on the way. Stay tuned for future developments. We will share more details when available.

Thank you, Andrew. This is David. I wanted to provide some additional insight on how I view this issue. We're discussing price and the total dollar amount associated with each vehicle. Ultimately, the determining factor for this dollar amount is the value it provides. This is rapidly evolving as we transition from Level 2 to Level 3. In the Level 2 phase, we consider items like airbags or seat belts, which are indeed critical for saving lives. However, while airbags save lives, there is an expected price point people have. For instance, no one would be willing to pay $10,000 for an airbag, even though it is life-saving. Therefore, we believe that a price below $200 is appropriate for such features, and we have aligned with that. This perspective is why we see such rapid growth in penetration rates; consumers recognize it as a good value. However, Level 3 is a different scenario. When contemplating Level 3, many people are already thinking ahead to Level 4. It’s important to consider value creation here. Aside from safety, the value for cars is the time that technology can reclaim for us. A Level 2 vehicle might provide you back one or two hours a day, but it doesn’t give much back because you still have to keep your hands on the steering wheel. This makes it a valuable safety device, but its overall utility is limited. In contrast, a Level 4 vehicle, like a robotaxi, could offer you up to 20 hours of use each day. A notable industry leader, who may not be a supporter of Hesai, once mentioned that the shift to this level could increase utility by about tenfold. If you can create a product that has the potential for ten times the traditional value, you can afford to invest in superior sensors and driving systems, ultimately allowing you to develop an exceptional product. That’s how I perceive Level 3 and especially Level 4. The dollar amount for a car might not necessarily be tenfold, but people are more willing to tolerate a higher price since it offers approximately ten times the value, if not more, compared to the Level 2 systems we previously had. This is why people should be excited about the overall content and the value created by a comprehensive sensor suite.

Your next question comes from Jesse Lo from Bank of America Securities.

Sorry, Operator, hold on for one second. Let me finish Jeff's second question regarding some additional color on our 2025 full year guidance. Jeff, as we said in the script, we are very confident that our full year numbers will fall in the range, but I'm not assuring that we will reach the high end of the range. So let's take the low end of the range, for example. We are basically guiding that our Q4 revenue will be above RMB 1 billion, comparing to roughly 800 million in Q3. If that achieves, the additional net profit comparing to Q3 will be roughly RMB 80 million pre-tax. If you look at our first 9 months results in this year, our total or accumulated net income is about 256 — sorry, it's USD 283 million. As we mentioned in the earnings release, we actually have a one-off investment gain from an equity investment, which is roughly RMB 150 million. So if we exclude that, our normalized first three quarters earnings is roughly RMB 130 million, so if we add another RMB 180 million to that, it already adds up to more than RMB 350 million net income in year 2025. So that's why we are relatively confident that our full year revenue and profit will fall in the range that we just mentioned. Okay. Operator, let's move on to the questions from BofA.

Could you provide some insight on BYD's 2026 LiDAR order since it is quite a contentious topic? Specifically, what is the timing for LiDAR adoption on this mass market model and what is Hesai's market share? Additionally, how should we approach the pricing strategy given the significantly higher vehicle sales?

Okay. We are pleased to see the leading domestic automakers actively accelerating their efforts to make intelligent driving mainstream. Their commitment exemplified by BYD's move to equip its models priced above RMB 100,000 with LiDAR and deployed its God's Eye ADAS systems. This creates massive market demand and accelerates consumer adoption, which benefits the entire ecosystem and raises awareness across the industry and consumers. With this trend, we are proud to be a key partner of BYD. We have gone into mass production with BYD since Q1 2025 and are supplying LiDARs for BYD's models launching in year 2025. This year, we are taking a strong share of BYD's LiDAR supply with our AT128P and ATX, with ATX leading in volume. Our partnership extends across double-digit vehicle models with an exciting wave of new SOPs rolling out through year 2025 and 2026. We will share more details on this customer, their new models and autonomous driving plans once they are ready to make an official announcement. Compared to automakers, Hesai has been investing in LiDAR R&D for 10 years, starting with more complex L4 applications. This has enabled us to accumulate extensive experiences and achieve superior product performances. Meanwhile, by leveraging our years of investments in ASIC technology, we have achieved excellent cost control while maximizing economies of scale through a broad customer base. We believe proactive collaboration between automakers and LiDAR companies helps create a win-win situation. In summary, Hesai serves as an index to the overall autonomous driving industry, and we are excited to see BYD leading the way in making intelligent driving more accessible while strengthening our partnership with them.

Your next question comes from Aaron Wang from Jefferies.

My question is about the Robotics side. Given the increasing demand in the robotic area in the coming years, could management share more color on our robotic shipments in 2026 and also in the next few years, and also the proportion of different end markets and our product mix in this segment?

Our Robotics business, which generally enjoys higher ASPs and margins is contributing strongly to the company's overall financials with the growth momentum accelerating. For the robotaxi part, Hesai is the largest robotaxi LiDAR supplier in the world, holding roughly 60% to 70% market share. Our main LiDAR and blind-spot products are widely used by all the top robotaxi players in China. Companies like WeRide, Baidu, Apollo Go, DiDi, Pony and Hello have all adopted our technology. Traditionally, robotaxi operators relied on mechanical spinning LiDARs for small fleet testing and operations. But recent trends in China show an urgent need for scalability. By using our flagship ADAS LiDARs, customers can achieve a better balance of sensor prices and performances, enabling faster fleet growth and helping them move closer to profitability. Spearheading this shift, we have recently signed new deals with WeRide, Pony and Hello Inc. Excitingly, for some of their models, up to 8 LiDAR units per vehicle, main LiDAR and blind-spot LiDARs will come entirely from Hesai. At the same time, we have signed new LiDAR supply agreements with leading global autonomous driving companies, including Motional across North America, Asia and Europe. These programs represent tens of millions of dollars in deals with strong follow-on potential as their partners scale up. The difference is that global players tend to favor high ASP mechanical spinning LiDARs for their performance and stability, making them less price sensitive. Regardless of LiDAR type, the gross profit from our robotaxi business is calculated as fleet size times number of LiDAR per vehicle times ASP times gross margin ratio. With cost reductions and large-scale commercial deployments, we expect the fleet sizes of leading players to grow exponentially in the coming year, driving significant profit growth for Hesai, whether they use mechanical or ADAS LiDARs. The other robotics applications beyond the robotaxi, we have seen huge opportunities in non-auto applications with intelligent robots rapidly gaining traction, our proven strength in high-performance, scalable LiDAR positions us to take the lead. Since starting production of our JT Robotics LiDAR, we shipped around 40,000 to 50,000 units every quarter in year 2025. This new product serves a wide range of robotics applications. Home, factory and agricultural robots are clear front runners, freeing people from routine work and creating real value from day one. In the long run, we believe the robotics market could have a TAM several times larger than ADAS. You can only drive one car, but in the future, 10 robots could be working alongside you. We anticipate that Robotics LiDAR volume could double in year 2026 versus 2025, and we plan to share updates on developments along the way.

Your next question comes from Chunsheng Xie from Huatai Securities.

My question is about the major customers for next year. Could you please share which OEMs will be the key customers for the ADAS products? And what kind of the demand scale or volume you are expecting for the next year?

In the ADAS spaces, we are also seeing very strong momentum from a number of key OEMs. Based on our current visibility, we expect the following OEMs to be among our top ADAS customers in year 2025 and very likely in year 2026 as well. They are Li Auto, Xiaomi, BYD, Leapmotor, Zeekr and Great Wall Motor. Of course, these names are not ranked in particular orders. Meanwhile, more major customers are also kicking off SOPs with us extending into year 2026, including Geely and Chery. These customers are rapidly advancing in intelligent driving with some of them expanding LiDAR adoption across more vehicle lines as a standard feature preparing for L3 capabilities with market LiDAR configurations. These companies are not only leading players within China's rapidly evolving smart EV sector, but also represent a diverse range of vehicle platforms from high-end to mass market vehicles, providing a rich foundation for our technology adoption and expansion.

Your next question comes from Sia Huang from SPDB International.

This is Sia. I've got just one question regarding our overseas update. And for the project with the top European OEM customer, is everything on track? And could you please also share more color about the overseas update in terms of other potential customers and design wins? And how do we expect the contributions of overseas revenue for the next 2 to 3 years?

Beyond our success and solid base in China, we are also stepping up our game internationally, both in ADAS and Robotics. On the Robotics side, recently, we have further strengthened our leadership with new LiDAR supply deals across North America, Asia and Europe, covering everything from robotaxi, robotrucks to robovans and factory automation. For example, we are proud to be the exclusive supplier for Motional's next-generation all-electric robotaxi. And as we shared earlier, we also signed a multi-year deal worth over USD 40 million with a leading U.S. robotaxi company, with deliveries running through 2026 and room for more as their fleet grows. In ADAS, momentum is just as strong. As you all know, we have already secured an exclusive design win with a top European OEM and several more are now in the sourcing and negotiation stage for their global programs with European players clearly setting the pace. As competition heats up, global OEMs are doubling down on autonomous driving, and safety is something they are never compromised on. There is now a clear consensus across the industry that LiDAR is becoming the airbag for autonomous driving, especially for L3 and above. And once these global OEMs make up their minds, they move decisively. So give them a little time to gear up their AD versions, both ICE and EV, and we believe more LiDAR deals will follow soon. Now for the global OEMs China JVs, things are also moving faster. They are right at the front line of the ADAS race. We have already won design win programs with five major JVs, Volkswagen, GM, Audi, Toyota and Ford and several are already in SOP. A recent highlight came in September when Audi E5 Sportback featuring Hesai LiDAR standard configuration hit the market and racked up over 10,000 orders in just 30 minutes. That's a huge commercial validation and sets the stage for global expansion ahead. Another exciting part of our strategy is backing Chinese OEMs in their global push. We will be the LiDAR supplier for models heading overseas with mass production kicking off in year 2026. We can't share more for now, but stay tuned. Updates are on the way. All in all, these wins reflect the growing trust and recognition we have earned from customers worldwide. Looking ahead, we will keep building on our strengths and serve an even broader range of partners across regions and across industries.

Your next question comes from Lou Jia from BOCI.

Congratulations on the excellent results. My question is related to our new business initiatives. At our Hong Kong listing event, management mentioned that for the next decade, Hesai will be more than just a LiDAR company. So could you share with us more about the potential new areas beyond the LiDAR, Hesai is considering expanding into in the future?

Thank you for the question, David. There are several exciting opportunities that we see ahead of us with significant potential and large total addressable markets where we believe we are well-positioned. First, we are still focused on sensing technology, particularly for robots and cars, with an emphasis on safety. Achieving safety is not about settling for 99% or 99.9% accuracy; it’s about identifying failure cases and continually improving products to enhance safety. We believe there are many ways to advance sensing technology to elevate safety standards, such as extending range, increasing density, recognizing various materials, and measuring speed. This presents a market much larger than the LiDAR total addressable market we currently recognize, considering the importance of these capabilities. Secondly, we’re integrating advanced sensing capabilities with artificial intelligence. We have developed a perception software stack and are collaborating with OEMs, although we are not yet charging them for this software, since the value we offer is part of their existing software stack. Many robotics applications and sensor uses show promise for integrating AI capabilities with our leading sensing hardware, enhancing both the ability of sensors to perceive and to process information. Thirdly, with software and AI advancements applied to top-tier sensing technology, we unlock possibilities beyond automotive applications, venturing into broader opportunities in spatial sensing. While I can't share all the specifics at this moment, we are observing strong demand from customers for technology that enables comprehensive sensing and 3D mapping of the environment. Imagine the capabilities of recording our 3D surroundings with our products. Lastly, while we’ve primarily discussed sensing, it’s important to note that our efforts in this area are foundational to our advancement into physical AI 1.0 for vehicles. As we move toward version 2.0, we’re not just developing LiDAR sensors for robots; we are expanding into a variety of infrastructure technologies where our established expertise in high-end sensing, semiconductor manufacturing, product iteration, and quality can be leveraged. In closing, I’d like to quote Jensen Huang, who once said that in the future, anything that moves will be autonomous. I believe anything that moves autonomously will need our advanced sensing capabilities. Thank you.

Thank you once again for joining us today. If you have any further questions, please feel free to contact our IR team. This concludes today's call, and we look forward to speaking to you again next quarter. Thank you, and goodbye.