XPeng Inc.

Hello, ladies and gentlemen. Thank you for standing by for the fourth quarter and fiscal year 2025 earnings conference call for XPANG Inc. At this time, all participants are in a listen-only mode. After management's remarks, there will be a question and answer session. Today's conference call is being recorded. I will now turn the call over to your host, Mr. Alex Gee, Head of Investor Relations and Capital Markets of the company. Please go ahead, Alex.

Thank you. Hello, everyone, and welcome to Expo's fourth quarter and fiscal year 2025, our new conference call. Our financial and opportunity results were issued by our newsletter service earlier today and available online. You can also view the press release by visiting the IR section of our website at ir.xiaofeng.com. Participants on today's call from our mentoring team will include co-founder, chairman, and CEO, Mr. He Xiaofeng, vice chairman and president, Dr. Brian Wu, Vice President, Mr. Charles Zhang, Vice President of Madison County, Mr. James Wu, and myself. Management will begin with prepared remarks and the call will conclude with a Q&A session. A webcast replay of this conference call will be available in the IR section of our website. Before we continue, please note that this discussion will contain forward-looking statements made under the safe harbor provisions of the U.S. Private Securities Litigation Reform Act of 1995. Four looking statements involve inherent risks and certainties. As such, the company's results may be materially different from the views expressed today. Further information regarding these and other risks and certainties is included in the relevant public filings of the company as filed with the U.S. Securities and Exchange Commission. The company does not assume any obligation to update any four looking statements except as required on the applicable law. It is also noted that EXPO's earnings price release and this conference call include the disclosure of unaudited GAAP financial measures as well as unaudited non-GAAP financial measures. EXPO's earnings price release contains a reconciliation of the unaudited non-GAAP measures to the unaudited GAAP measures. I will now turn the call over to our co-founder, chairman, and CEO, Mr. He Xiaopeng. Please go ahead.

Hello, everyone. Back in 2015, many areas of Xiaopeng has achieved a return-based breakthrough. In 2015, our annual turnover was 429,445 units, which is 126% of the total growth. Among them, M03 became is the current champion of the 150,000-level pure electric car, and PG-JIA is the number one pure electric car in the 150,000- to 200,000-level pure electric car category. Our Kunpeng Super-Extra has achieved mass production on the X9, and opened up the era of Xiao Peng's one-car dual-energy. The quality management system of our company and the user satisfaction of NPS are increasing rapidly throughout the year. In 2025, we will be able to sales nearly doubled to 4.5 million units, and overseas revenue accounted for more than 15%. Not only that, but also thanks to the upgrade of physical AI, we achieved the mass production of Turing AI chip and handed it over to our partner, the public. The second generation of VIA also came to an eternal moment. The appearance of Ion-type machines has also aroused global attention. In 2025, we will not only We actively responded to the national call to shorten the supply chain period by 50 days. On this basis, the free cash flow of the whole year is close to 5 billion yuan. At the end of the year, the debt cash is 477 billion yuan. Hello everyone.

Reflecting on 2025, XPang delivered a series of milestone breakthroughs across multiple fronts. Our annual delivery reached 429,449, up 126% year-over-year. Mona M03 became the best-selling battery electric sedan in the 100,000 to 200,000 RMB segment. B7 Plus ranked number one among pure electric sedans. in the 150,000 to 300,000 RMB segments. The Kunpeng Super Extended Range EV X9 entered mass production and started our one vehicle dual energy era. Our quality management system and customer satisfaction, as measured by NPS, improved rapidly. Overseas deliveries nearly doubled to 45,000 units, with revenue from overseas markets contributing over 15% of total revenue. In addition, Benefiting from enhanced organizational capabilities centered on physical AI, we successfully brought our Turing AI SOCs into mass production and began deliveries to Volkswagen. Our VLA 2.0 saw an emergence, and the debut of our humanoid robot, Iron, garnered significant attention globally. In 2025, we achieved not only high growth in scale, but also significant improvement of operational capabilities. Our gross margin for fiscal year 2025 reached 18.9%, an increase of 4.6 percentage points year over year. In response to national initiatives, we substantially reduced the turnover days of accounts available to suppliers by 60 days. We had a free cash flow inflow of approximately 5 billion RMB in 2025 and ended the year with 47.7 billion RMB cash on hand, providing robust financial support for unwavering investment in physical AI RMB. In Q4, Xpeng achieved its first ever quarterly profit with net profit exceeding 380 million RMB, This Moxie imagines a business model driven by technology leadership, which is a path to profitability distinct from that of traditional automakers.

我特别开心的是通过多年的硬件规模研发投入, 并且在最近数年我们对AI的真正的转型。 我相信小鹏汽车正处在物理AI应用的关键转折点。 我们构建的芯片, machine modeling, data, EEA architecture, and AI infrastructure, including technology and organization. We are seeing that autonomous driving will become the first physical AI that can be widely used and has a huge commercial potential. It is also in the process of integrating with the AI. It is also in the process of integrating with the AI. I believe that we will continue to break through the upper limit of technology

I'm glad to see that with years of hardware investment and physical AI transformation, XPang is at a pivotal inflection point in the application of physical AI. After years of sustained and significant R&D investment, we have built a full stack in-house developed technology system, including SOC, foundation model, data, W-E architecture, and AI infrastructure. We're now driving upward integration across autonomous driving and smart corporate systems, while enabling cross-domain integration downward into power chain and chassis systems. Looking ahead, we'll continue to push the technological frontier, accelerate the development of AI, XBI deployment, and commercialization of our product technology innovations.

At the beginning of this month, Our second generation VA has passed the physical test of autonomous driving. When a passenger is already difficult to distinguish whether the driver is driving or the AI is driving, the car is gradually becoming an agent of the physical AI. I believe that Xiaopeng's second generation VA will soon cross the threshold of technology and scale applications, and complete autonomous driving will arrive in the next one to three years. In the near future, AI will bring huge changes to everyone's travel, work, and life.

In early March 2026, VLA 2.0 successfully passed the physical Turing test for autonomous driving. Passengers can hardly distinguish if it's a human or AI who's driving the car. The vehicle has effectively evolved into a physical AI agent. XPeng's VLA 2.0 will soon cross the inflection point of both technology and large-scale deployment. Fully autonomous driving can be expected to come in the next one to three years. AI-powered vehicle and humanoid robots will soon fundamentally reshape how everyone travels, works, and lives.

我相信今年是小鹏的产品线扩容的大年, 同时也是我们的产品力, we plan to launch four new models from large to compact models. These models are designed by the new group of re-buyers and are developed in the global market. They all have one-car dual-energy and support automatic driving capabilities, including L2 and L4. I believe that with the continuous delivery of new models, I am confident that sales

2026 marks a major year of product portfolio expansion and upgrades in our product capabilities. We plan to launch four new models, expanding into both large and compact size vehicles. Designed by our newly restructured design team, these models are crafted for global markets and built on dual energy platforms, supporting the evolution of autonomous driving capabilities From L2 Plus to L4, as deliveries of new models ramp up, we'll achieve strong quarter-over-quarter growth in volume.

In the second quarter of this year, our first flagship and sixth-generation SUV, Xiaotong GX, will be officially launched. It will bring a family user an experience that does not need compromise. GX not only has a comfortable large-space MPV level, but it will also carry a live-action It will also be the first car model that supports L4-level hardware capabilities. So from this year on, Xiaofeng will not only expand the global market share of AI cars, but also let the second generation of VIAs go to the world. I will also lead the team to realize RoboTaxi, which can drive unmanned vehicles, flying cars and high-speed vehicles.

In Q2 2026, we'll start pre-orders of our first flagship six-seat full-size SUV, the X-Pan GX, bringing in a truly uncompromised, best-in-class experience model to family users. The GX delivers MPV-class comfort and spaciousness, and features nearby wire and rear-wheel steering. It'll be our first model designed to support L4-level hardware and software capabilities. Starting from this year, we'll further expand our AI vehicles' global market share and bring our VLA 2.0 model to global markets. At the same time, I'll lead teams to bring both RoboTaxi and advanced humanoid robots into mass production.

After the global supply of products, in 2026, I hope to make the global supply of mass-produced robots more solid. We will continue to make mass-produced robots in many aspects and continue to improve the quality and reputation of internationalization. Our goal is to increase overseas sales by threefold this year. So overseas business income contributions have increased by more than 20%. We have four new cars in the international market this year to build a global star model in an important SUV category. Our overseas sales target this year is to reach 680 stores. will be doubled by 25 points, and Xiaofeng's self-sustaining control network will cover 10 overseas core markets from China to the world, bringing new generations of overseas users with 5C super power. I believe that based on the ability to improve the productivity of product technology and production and sales in the entire overseas system, the internationalization process of Xiaofeng will be further improved in 2027 to 2028.

With global product sales increasing, in 2026, we'll further solidify our global scale-up capabilities across production, supply, sales, and service. Our supply chain, manufacturing, logistics, spare parts, sales, and service systems are moving towards greater global synergies, underpinning sustained improvement in quality and brand perception in international markets. For our overseas deliveries, the goal is set to double in 2026 year over year, with international business contributing over 20% of total revenue. We plan to introduce four new models to global market. In the key SUV segment, we aim to launch more global flexion models. We target to have 680 overseas stores for sales and service, doubling the number of stores from the end of 2025. Xpeng's self-operated ultra-fast charging network will also expand beyond China to 10 key international markets. This will bring our industry-leading 5C ultra-fast charging experience to users globally. With an enhanced overseas competencies, our global expansion will further accelerate significantly through 2027 and 2028, and revenue from overseas markets will become one of the core drivers of the company's profitability.

In order to cross from L2 to L4, and to achieve global deployment, in the past year, we have done a great job of redevelopment of the entire technical structure of Autopilot, as well as the organization. Since yesterday, our second-generation VR model has officially launched for user continuous promotion. I believe this is the first version of Autopilot delivered to us. It is safe. We hope to become a mother who loves to open the national brand. We have already opened 732 stores in the country. After the opening, our Japanese army in March increased by one times the amount of four stores. To make the leap from L2 Plus to L4 and enable global deployment, we have fundamentally redesigned the architecture of our autonomous driving technology.

Starting yesterday, we began the gradual rollout of our VLA 2.0 to users. This is our first release designed for the L4 fully autonomous driving era, delivering a safe and smooth experience with driving performance comparable to experienced human drivers and representing a generational gap ahead of the industry. The VLA 2.0 is transforming advanced autonomous driving from an early adopter feature into a truly mainstream mass market feature that even everyday drivers can trust, rely on, and enjoy with confidence. All of our 732 stores nationwide now offer VLA 2.0 test drives. So far in March, the number of our daily test drives has doubled month over month, and the percentage of the Ultra and Ultra SE trims among all trims has also more than doubled. Over the next three to six months, VLA 2.0 will notably improve test drive conversion rate and substantially increase user engagement and retention for XPeng.

The second generation of VLA models basically broke all the rules and verified the scaling in the physical world. Next, we will firmly expand it We believe that the average acceptance rate will increase by 25 times compared to the industry, and the safety acceptance rate will increase by 50 times. According to this speed of improvement, I think the full automatic value will arrive in the next one to three years. It will make autonomous driving a daily habit for people. With the second generation of VL models, the RoboTaxi model, we have already obtained the license for the Guangzhou smart network car road test today, and we are currently conducting a long-term L4 public road test. We are planning to open the RoboTaxi in the second half of this year, and we will also conduct a test in the overseas market. The VLA 2.0 has broken down traditional rules and validated the applicability of scaling laws in the physical world.

Next, we'll continue to scale up to aggressively widen our leads. By the end of the year, we target to increase the number of parameters on the edge from several billion to 20 billion level. This will increase average miles per takeover by 25 times and the safety critical miles per takeover by 60 times. At this pace of improvement, the era of fully autonomous driving is on track to emerge within the next one to three years. Autonomous driving will become part of everyday mobility. Our GX RoboTaxi, powered by Reality 2.0, has received official road testing approval in Guangzhou and is now conducting ongoing L4 public road tests. In the second half of the year, we plan to launch pilot passenger operations for our RoboTaxi service to validate the technology, user experience, and the business model. We'll also begin overseas road testing of the VLA 2.0 as technology advances in red regulatory frameworks evolved, XPeng will be one of the few companies globally that is capable of scaling autonomous driving efficiently across multiple global markets.

第二代VI模型能够有如此强劲的性能跟效果, 得益于小鹏自研图灵芯片, 并且联合优化了自研的基模跟编译器。 我们原计划,自研的芯片是有商的三倍有效权利, 但实际上, has achieved a 10-fold improvement. Since the production of Turing chips in the third quarter of last year, the total output has exceeded 200,000 chips. This year, all full-size models from the second quarter, including MaxLemon, will be completed and switched to Zhiyuan's Turing chips. This year, Zhiyuan's chips' annual output target is about 1 million chips, and it is expected to become the No. 1 in the output of China's large-capacity Dante AI chips. Then the general public is our Turing chip with the second generation of VIA outside cooperation of the first customer with the deep cooperation and high efficiency production of the top-level car companies. So it confirms that our technology has a universal system that can be used and exported. We welcome more car companies and smart enterprises and TEL1 to use our Turing chip and the overall solution of the automation to serve customers.

The strong performance of our VLA 2.0 is underpinned by the joint optimization of our in-house Turing SOC, proprietary large foundational large model and compiler, delivering a 10 times improvement in effective compute, even though the plan was only to achieve a three times improvement. Since entering mass production and being deployed in vehicles in Q3 last year, the Xpeng Turing SOC has actually shipped over 200,000 units, starting from the second quarter of this year. All Xpeng models, including the MAX trims, will fully transition to our in-house Turing SoC shipment. Shipments of Turing SoC are targeted to reach nearly 1 million units this year. The Turing SoC is positioned to become the leading high-compute AI SoC deployed on the edge by shipment volume. Volkswagen is our first external customer for both our Turing SoC and VLA 2.0. Our deep collaboration with leading global automakers, together with our ability to mass produce efficiently, demonstrates that our technology is scalable, replicable, and globally competitive. We welcome more automakers, embodied AI companies, and Tier 1 suppliers to adopt our Turing SoC and integrate intelligent solutions for their customers.

In my opinion, the competition between AI and physics is more and more due to the change in the efficiency of technology, and it ultimately depends on the compatibility between the underlying organization and AI. In recent years, we have completed a very critical organizational upgrade. We have established a comprehensive artificial intelligence center with an integrated artificial intelligence center. For the overall development of cars and robots, I believe this is not only a structural adjustment, but also a development of the behavior of the model. So from now on, no matter whether it is a machine or not, decision-making, interaction, the elders and the elders will not be divided. How do we drive on the road? How do we communicate with the car? We will share a few models and sound methods. The efficiency and efficiency of the development has been greatly improved. So what I am very happy about is that in the past few years, Xiaofeng has run through the entire chain of sound methods, from end-to-end computing, cloud computing, data editing, quantification, deployment, world verification, In my view,

The ultimate competition in physical AI will be determined by fundamental organizational capabilities and AI infrastructure. Recently, we completed a critical organizational upgrade, integrating our autonomous driving center and smart cabin center into general intelligence center. This is more than a structural change. It represents a paradigm shift in building intelligent systems, driving systems decisions, and human-vehicle interactions will no longer exist in isolation. How the vehicle drives on the road and how humans interact with it now share the same physical AI foundation model and infrastructure. R&D efficiency and effectiveness have improved substantially. Xpeng has established a full integrated AI infrastructure stack, including computing power on the cloud and on the edge, data compilation, quantization, deployment, simulation, and real-world testing in great scale. Our physical AI foundation model will evolve at an exponential speed. It will power vehicles, mobile taxis, and humanoid robots and become a unified super foundation model for all of our physical AI agents.

同时我们的目标在今年年底我们量产我们的Iron汽车。 小鹏Iron汽车 with three Turing-A chips. We have the most efficient and far-reaching infrastructure in the industry. Our current technical route is different from that of many other infrastructure companies. Not only do we do a good job of our own, but we are also very focused on how to make the big brain responsible for self-reflection and decision-making, the small brain responsible for movement, and the entire control system, as well as the data accumulation of the giant intelligence model. What makes us particularly happy is that our latest VLA2.0 technology station has successfully run on robots. Combined with Xiaofeng Robotics' new fourth generation exercise control system, I believe we will see the best exercise and control effect in the industry in the next few years. I am also very confident that Elon's intelligence level will open the way for the industry to expand. Xiaofeng Robotics will take the lead in business in order. Thank you. Our next generation humanoid robot IRON is targeted to enter mass production by the end of 2026.

Powered by three Turing AI SoCs, IRON's computing power on the edge far outpaces most robots in the industry. Our technology focuses on the robot's brain and cerebellum. The brain enables independent thinking and decision-making, while the cerebellum controls motion. At the same time, we continuously accumulate data for embodied intelligence models. Our VLA 2.0 technology SAC is now running successfully on our robots. Paired with the fourth-generation motion control system, IRON will deliver industry-leading agility and motion control in the second half of the year. establishing a clear generational lead in intelligent capabilities. RN will focus on three key application scenarios, commercial, industrial, and household. Initial deployment will support reception, guidance, and retail assistance across Xpeng stores and campuses in China and overseas. In the first quarter of the year, we began construction of our humanoid robots mass production base in Guangzhou. RN aims for a monthly production target of over 1,000 units by the end of this year. Leveraging Iron's advanced intelligence and our strength in mass production with quality and supply chain management, XPeng will become one of the world's largest and most valuable humanoid robot companies.

Because we have already seen the physical AI world, This is a huge downfall of the original software development. We will continue to move very firmly on the development path. So in 2025, our Xiaopeng's development investment is 9.5 billion yuan. Among them, the investment in AI reached 4.5 billion yuan. So in the past few years, we have built a strong and efficient investment in the AI field. We have built the industry-leading physics and AI. These two fields are integrated into the whole war system. such as chip, Gmod, and Infra. I believe that with the foundation of this system, in the beginning of the year of 2026, the integration of AI technology and the performance of the business will be greatly accelerated. Then, the A-level mass production of our AI will be used to open up the upcoming growth curve. I believe that in the next five to ten years, the application market space of AI will be wider than that of the automotive industry, We believe that the global RoboTaxi and the energy generation are all in the 1 billion to 1 billion level market. At the same time, we believe that the agent sales list of physical AI, including physical AI cars, will be similar to the sales list of gasoline cars 10 years ago, and will be a very rapid change like today's new energy sales list.

Because we already saw the tremendous disruption that the physical AI world has brought to the original software development paradigm. In 2025, we invested 9.5 billion RMB in R&D, including 4.5 billion in AI. Our sustained and efficient investment in AI R&D over the past few years has enabled us to build an industry-leading, fully in-house physical AI technology stack, including SOCs, foundation models, and AI infrastructure. Powered by this technology stack, the pace of advancement in physical AI will accelerate significantly starting 2026. We'll see the mass production and application of physical AI agents of their scale poised to enter a steep growth curve. Over the next five to 10 years, the market for physical AI is expected to surpass that of the automotive industry. Those Robotexy and humanoid robots represent trillion to $10 trillion level global market opportunities. In the future, the sales rankings of physical AI agents may matter even more than the ICE vehicle volume chart a decade ago or any V volume chart today and they will change very, very fast. We're confident that XPeng will become the global leader of physical AI agents.

We are more determined to increase the development intensity of AI and the physical world than ever before. In addition to policy development, The investment in AI-related research and development of physical AI will increase to 71 billion yuan. In my opinion, the development of AI not only has a long-term improvement, but will also create huge commercial returns in the short, medium, and long term. Then the expansion will allow us to win, survive, and win in the middle of competition. Then we will build our new core economy in the lead of physical AI technology and commercialization.

Pushing the boundaries of physical AI is very exciting for my team and me. We're more committed than ever to intensifying our R&D investment. This year, in addition to vehicle development, investment in physical AI-related R&D will increase to 7 billion RMB. In my view, this investment will not only help us increase our competitive advantage and deliver substantial long-term returns, scale allows us to survive in competition. but it is sustained leadership in physical AI technology and commercialization that will define our core competitive advantage.

I am very happy that our market is recovering quickly after the Spring Festival. I expect the delivery volume in March to increase by 69% to 101%. With the delivery of the second-generation VLA and four new car models, I believe that this year's subsequent quarter sales will rise sharply, and in the second half of the year, it will achieve the same growth as the long-term industry. We are right on the front page of the size of the physical AI system. In the future, I will use more energy to build a globalized Xiaohong, business strategy, organize and operate the system, and turn AI and physics into real business success. I think that AI will innovate a new business model. It will increase the size and effect of the original car, as well as the new network effect and agent effect. The record of the market will be concentrated and greatly improved. As the leader of physics A, we will have a longer-term leading edge.

For the first quarter of 2026, we expect deliveries to be between 61,000 and 66,000 units. Revenue is expected to be between 12.2 billion and 13.2 billion RMB. March deliveries are expected to grow sequentially by 69% to 101% month over month. As the VLA2.0 and four new models enter mass production, we expect quarterly sales to continuously trend upward and achieve year-over-year growth in the second half of the year to significantly outpace the industry. We are on the cusp of large-scale mass production for physical AI agents. Going forward, I will dedicate more of my efforts to develop strategies, establish organization and operational capabilities for Xpeng's globalization and commercialization, with a focus on translating our technological leadership into commercial success. The innovative business models of physical AI will add network effects and agent effects on top of the traditional automotive economies of scale. meaningfully raising market entry barriers and industry concentration. As a leader in physical AI, we'll possess a long-term sustainable advantage, securing a higher share in a vastly expanded market and achieving greater corporate value.

谢谢大家的聆听,下面由我们的财务VP James为大家介绍我们2025年第四季度的财务状况。

Thank you, everyone. With that, I'll now turn the call over to our VP of Finance, James, who will walk you through our financial performance for the fourth quarter of 2025.

Thank you, Xiaofeng. Now, let me provide a brief overview of our financial results for the fourth quarter of 2025. I'll reference RMB only in my discussion today, unless otherwise stated. Our total revenues were $22.25 billion for the fourth quarter of 2025, an increase of 38.2% year-over-year and an increase of 9.2% quarter-over-quarter. Revenues from vehicle sales were $19.07 billion for the fourth quarter of 2025, an increase of 30% year-over-year and an increase of 5.6% quarter-over-quarter. The year-over-year and quarter-over-quarter increases were mainly attributable to higher deliveries. Revenues from services and others were $3.18 billion for the fourth quarter of 2025, representing an increase of 121.9% year-over-year and an increase of 36.7% quarter-over-quarter. The year-over-year and quarter-over-quarter increases were primarily attributable to the increased revenues from, firstly, the technical R&D services rendered to the Volkswagen Group due to the successful achievement of certain key milestones in the current quarter. Secondly, parts and accessory sales in line with higher accumulated vehicle sales. And lastly, carbon credit trading. Gross margin was 21.3% for the fourth quarter of 2025, compared with 14.4% for the same period of 2024 and 20.1% for the third quarter of 2025. Vehicle margin was 13% for the fourth quarter of 2025, compared with 10% for the same period of 2024 and 13.1% for the third quarter of 2025. The year-over-year increase was primarily attributable to the ongoing cost reduction and improvement in product mix of models. R&D expenses were $2.87 billion for the fourth quarter of 2025, representing an increase of 43.2% year-over-year and an increase of 18.3% quarter-over-quarter. The year-over-year and quarter-over-quarter increases were mainly due to higher expenses related to the development of new vehicle models and technologies as the company expanded its product portfolio to support the future growth. SG&A expenses were $2.79 billion for the fourth quarter of 2025, representing an increase of 22.7% year-over-year and an increase of 12% quarter-over-quarter. The year-over-year and quarter-over-quarter increases were primarily due to higher commission to the franchise stores related to the sales volume and the launch of new models. The year-over-year increase was further due to higher marketing and advertising expenses. As a result of the foregoing, loss from operations was $0.04 billion for the first quarter of 2025 compared with $1.56 billion year-over-year and $0.75 billion quarter-over-quarter. Net profit was 0.38 billion for the fourth quarter of 2025, compared with net loss of 1.33 billion year-over-year and net loss of 0.38 billion quarter-over-quarter. The company recorded positive net profit for a single quarter for the first time in Q4 of 2025. As of December 31, 2025, our cash position was $47.66 billion. To be mindful of the length of the earnings call, I would encourage listeners to refer to our earnings press release for more details on our fourth quarter full year 2025 financial results. This concludes our prepared remarks. We'll now open the call to questions. Operator, please go ahead.

Thank you. If you would like to ask a question, please press star 1 on your telephone and wait for your name to be announced. If you would like to cancel your request, please press star 2. If you are on a speakerphone, please pick up the handset to ask your question. For the benefit of all participants on today's call, if you would like to ask your question to management in Chinese, please immediately repeat your question in English. For the sake of clarity and order, please ask one question at a time, management will respond, and then feel free to follow up with your next question. Your first question today comes from Tim Shao from Morgan Stanley.

Please go ahead.

Hello, Manager Chen. I'm Morgan Stanley of TAM.

Thank you for accepting my question. Congratulations to the company on achieving the profit of mobile phones in the fourth quarter. I have two questions related to the price. The first question is, I want to know what kind of upgrade plans will be available in the next year of Xiaopeng VOA 2.0. I expect Xiaopeng VOA 2.0 I have two questions that both are related to smart driving. The first one, what type of major upgrades should we expect for X-Pen's VLA 2.0 in the coming months? How do you anticipate VLA 2.0 will impact order conversion? and the user retention in the following quarters? That's my first question.

Thank you. Let me go back to your first question. In the following each quarter, we have at least one large-scale OTA. I think we can share some of it. For example, in the second quarter, we will include This self-driving car only covers major roads in high-speed urban areas. We will start to cover more small roads, parking lots, and roundabouts. In fact, this is a very critical change. That is to say, from a self-driving car with a public road with navigation, it has become a self-driving car with no navigation and no rules on the internal road. In fact, in the robot system, basically all of them are like this. There are four types of self-driving cars and navigation. Then we will do a lot of This is an ability upgrade. For example, we will upgrade the car model from the 1 billion level to more than the 2 billion level. We hope that the safe access process can be increased by 5 to 10 times in this year's modern market. We actually have a lot of upgrades other than the automatic value, such as adding multi-language support in localization, such as increasing the integration of more factory and automatic value. Let me turn over the first paragraph. Thank you.

Thank you for your question. Let me answer your first question first. So basically going forward, each quarter we are expecting at least one major version of OTA. And I think we can share some of it just to give you some examples. In our Q2 OTA, for example, the autonomous driving will be able to actually cover more roads, whereas currently or traditionally it was only able to cover the major highways. And in the future, for Q2 OTA, we'll be able to cover smaller parking lots and campuses or communities in different parks. It's actually a very critical change, meaning that our capability is evolving from navigation-enabled public roads to cover more areas. In addition to that, we would like to also highlight that autonomous driving in the future will also become more AI-agent-oriented. Actually, if you looked at all of the scenarios of applications for robot systems, essentially very similar to that, Also, the four levels of autonomous driving and navigation also follow the same logic as well. This year, we also expect to do a lot more upgrades in terms of our capabilities. For example, if you look at our model on the vehicle or the edge side, we are increasing the parameters from billion level to actually up to 20 billion level. We also hope that we can increase our mileage per takeover by about 5 to 10 times. In addition to that, apart from autonomous capability driving, I mean, upgrade, we also have added some multi-language support capabilities as well. So that's another level of upgrade for localization support, which will allow us to combine smart carpet with autonomous driving. Thank you.

Because our VIA2.0 has been gradually promoted to users since yesterday, I can say that since the beginning of March, We see that the capacity of the door shop is doubled. We see that the sales volume of the Ultra and Ultra SE versions has increased by more than one times. We believe that when the full volume is launched, and it is constantly upgraded, the second VIA will have, I hope, a higher customer flow and a further large-scale transformation, and it will also raise the price of all our single-car vehicles. So when it comes to the use of connectivity, I can only share a case of my own. After using the second generation of VLA, I will use another version of autonomous driving. Or because I often go on business trips, I need a lot of drivers to help me drive. I found that I already feel that no matter what other types of autonomous driving or professional driver driving, it can't compare to VLA's smoothness and peace of mind. So now I see that when I get used to VLA, maybe after two or three months of driving, I have completely changed it. I think high-level autonomous driving will definitely become 99% to 100% of the needs of daily life. I think this is a very, very strong feeling. My driver has asked me twice recently. From my point of view, when will this professional driver start to eliminate the problem?

Now, because our VLA 2.0 only started to be officially pushed to our users yesterday, and I mentioned in the prepared remarks that we started promoting it in early March. And as a result, we could actually see that the market really acted very, very positively. For example, our test drives actually more than doubled sequentially, you know, across our stores. And also, Ultra and Ultra S1's sales are also more than double as well. And I believe when we fully launch our VLA 2.0 to the market and to our users and continue to upgrade it, we are expecting actually a higher sales volume and also a higher level of conversion rate, which will also eventually raise the average selling price of our vehicles. Now, when it comes to user retention or engagement, I would like to share my own case and experience as well. After trying out the VLA 2.0, there's no going back for me, honestly, because I travel a lot, and a lot of the times I actually rely on professional drivers who drive me around. And after trying VLA 2.0, I realized when I now use other types of autonomous driving versions or, you know, when I ride our professional driver's cars, they're not nearly as good as VLA 2.0 because VLA 2.0 now gives me the smoothness of driving and also on top of that, peace of mind as well. That kind of experience that I've got offered is unparalleled. And it's been, what, only two, three months? And I'm no longer the same person. I mean, there's no going back, really. So I believe that in the future, advanced autonomous driving will definitely be part of everyone's daily necessity. User rate potentially will reach basically 100%. And I feel this very, very strongly. And recently, my professional drivers have been asking me a couple of times that, you know, from my perspective, when will our profession, you know, he was talking about being a professional driver,

be eventually phased out.

Now, I would like to just add one brief last point, which is that this is really representing a completely new paradigm, which from my perspective, that our current priority is to really address a lot of safety issues and many of our witnesses first. However, since the launch of our second generation VLA, you know, the rules have completely changed. It's never the same again. So I believe that with making up of our weaknesses or the weaknesses being compensated for, now we can actually spend a great deal of our time on enhancing our overall strengths and deployment globally. Thank you.

Thank you, Mr. Xiaopeng, for sharing your thoughts.

I would like to follow up on another question. Just now, Mr. Guan mentioned in the briefing that one of the key focus this year is overseas. I would like to know about Xiaopeng's Ultra model and ULA 2.0's deployment plan in the overseas market. What impact will it have on overseas sales? In addition, do you think this software will have the opportunity to receive independent fees overseas? My second question is also related to smart driving. So what is the deployment roadmap for X-Pen Ultra model and VOA 2.0 in the overseas market? And how will the overseas expansion of smart driving affect X-Pen's global sales? And does the smart driving software have the potential to be monetized alone? That's my second question. Thank you.

Let me answer your question briefly. First, we have already started the development and preparation work of the second generation of VIA. We expect that at the end of this year and next year, the second generation of VIA will gradually open up in different areas. I think that VIA has a very big advantage overseas. First, we found that it is very well developed. We found that in many market tests, There is no overseas data. Just rely on our VLA, we have already been able to achieve very good results. The second point, compared to our friend FSD, we are particularly good at small roads. In Europe, in Southeast Asia, our advantage is particularly obvious. So I was very confident last time that VLA can provide high quality and high security in overseas at a faster, lower cost and better. Thank you.

Let me briefly address this question. First of all, we have already launched preparation for testing our VLA 2.0. So by end of this year or beginning of next year, we are going to gradually roll out the VLA second-generation testing grayscale and also its delivery or deployment across different regions. And I believe that VLA 2.0 has a significant advantage overseas. First of all, we found that it's actually very generalizable. In our testing across multiple markets, we realized that even without any overseas actual road data, our VLA alone performed really well and achieved excellent results. The second point is that compared to some of our competition, for example, SSE out there, our VLA has a particular, you know, advantage over them in the competition. capabilities across smaller roads and also country roads or uncharted territories. And we've tested, we've seen that in some overseas markets, such as Southeast Asia and Europe. So that made me, again, you know, having a stronger conviction that our VLA 2.0 can actually provide high quality, safer, autonomous driving for our users overseas at a higher quality and also lower cost. So with that, we now have done all the preparation in the hardware. And so by second half of this year, we definitely are ready for a bigger test and also for the future launch as well. And I believe that we are also considering some business model upgrade or transition for our smart driving software in the overseas market. And I believe there's definitely a lot of opportunities for commercialization or monetization. And we have strong confidence in converting our great technology into good business opportunities and revenues.

Thank you. Your next question comes from Nick Lay from J.P.

Morgan. Please go ahead.

Okay, thank you, Mr. Peng and Mr. Wang. I'm J.P. Morgan, Nick. I have two questions. I also want to discuss the robot part, which is what I just mentioned, the self-driving part. The first part is related to the robot part. Mr. Peng just mentioned that the car exhibition in April will be held at the end of the year, etc., which is the mass production, etc. China China China China China My first question is really related to humanoid robot ambition and the long-term strategy. First of all, mentioned that in the near term, we'll see the product launch at Beijing Auto Show followed by mass production by year end. Aside from that, is there any major milestone that investors should be mindful in the next, say, one or two years? In addition to that, can we expand our footprint from a smart vehicle to a humanoid robot? Anything we can leverage regarding R&D, production, and other supply chain? Thank you.

Yes, thank you, Nick. I think, first of all, there is a part of the content of your question. We didn't say a milestone in April this year. We said at the end of the second half of this year. Your question is very big. I think, first of all, let's go back to a question. Last year, many entrepreneurs and supply chain partners asked us the same question. They said, is the supply chain of the robot a commercial one or a car one? We often think that it is at least a car cabinet. For example, a car has only one engine. If your engine is broken, then you are definitely an A-class driver. But a robot has 70 to 80 joints. A lot of joints are damaged, and a lot of signal loss may be a serious safety problem, if you look at it from the perspective of a car. So we started to develop in a car-oriented way in early last year. to supply chain locations, to prepare for production and manufacturing. So the overall logic, you will see that Xiao Peng is a very rational robot. He is using a car to go straight from joints to battery pack, from chip to machine, from data to the whole old man, little old man, big old man training. I think there will be a lot of difference. I think So in another perspective, we think we are using a robot that is more than ten times more difficult to achieve high-quality mass production. The goal of the end of this year is to reach 1,000 units. So the future of cars will be the company of the future. But the big difference between robots and cars is that the value generated by robots in software is at least 50% at the beginning. . . . . Thank you.

First of all, thank you for your question. And I would like to clarify something. We didn't mention April, but very much in the second half of this year. I just want to clarify that from your question. And again, you know, what you asked is very, very big, and I'll try to address it. First of all, Last year, a lot of supply chain partners and also a lot of companies asked us the same question about robot supply chain or humanoid robot supply chain and whether or not it should be car grade. We believe that car grade is the minimum requirement because, obviously, cars, they typically have one engine, and if the engine breaks down, it already caused A-class accidents. And when you look at humanoid robots, they have at least 70 to 80 joints, and any kind of damage to those joints will cause information loss or signal loss and will cause very, very serious accidents and will compromise our safety. And so we need to look at humanoid robots from at least a car grade level standards. And since beginning of last year, we already started the development for our humanoid robots adopting car grade standards as well in preparation for its manufacture. So overall, you can see that XPeng is a very unique robotic company in that sense because we are adopting the same kind of high-level rigorous of developing a vehicle in developing our humanoid robots. And you can actually see that we have to do a lot of in-house development covering not only the joints, but also the torso and other parts of the body from the SOC to the foundational model, from data to all the training of the brain and cerebellum, etc. And we are very unique in that sense. I believe that we are actually using a very, very challenging way or a very new methodology, which can be 10 times harder than developing RoboTaxi in doing our humanoid robots to achieve the highest quality mass production. And by end of the year, our target of production is at least 1,000 units. And I think future vehicle companies will also be robotic companies. However, one of the biggest differences between humanoid robots and vehicles is that robots actually generate a lot of value from the software, and it can start at 50% already since day one, and cars may start slow, you know, with only 10%, 15%. It will take them some time to actually ramp up in the value creations. But I believe that they actually come from the same logic and also share the same origin. So whether when it comes to R&D, mass production, supply chain management, quality, etc., and also the global management, globalization management, the supply chain sales, marketing services, etc., there's going to be a lot of similarity between car and robotic development. And that sets us apart from our competition because we have always been doing things in-house with full stack technology and also cross-domain integration. And fundamentally, you know, we use the same kind of logic as well. When we develop our robots from the, I would say, elbows, hands to legs to feet, you know, we do everything in-house or at least we do joint development, which, again, set us apart from other car companies or robot companies. Thank you.

My second question is also related to human robot regarding the cost and the selling price. And how should we expect the cost reduction after mass production from year end? And also, should we expect any external sales after mass production from year end? Thank you.

This is a very good question. I think the cost logic of the robot is different from the cost logic of the car. Because the cost of the robot, I think there are three costs. The first is hardware cost. The second is R&D cost. The third is operation cost. So, I think hardware costs may be close to cars, but the cost of research and development related to AI will be much higher than that of cars, especially in the past. Car habits and T1 cooperation can bring other people's software and hardware solutions over. That's in the rule of thumb. But in the age of AI, the same rules are brought over. The same AI is not a rule. It can't be brought over directly like this. Then the third is the operating cost, because you will constantly have new data . . . . but the cost of software and operation will continue to increase. What are the application scenarios for machine production? We are different from other companies. First of all, we are not facing the research market. At the beginning, we wanted to enter the business, then enter the industry, and then enter the family. These three scenarios depend on our changes in technical capabilities. Entering the business requires a lot of safety. It requires a lot of mobility. Thank you.

This is a very good question. And I think when it comes to, you know, thinking about the cost structure for humanoid robots, first of all, it's very different from what we've seen in cars. There are three major cost items for a humanoid robot. First of all, you have the hardware cost and then the R&D cost. Lastly, you have the operational cost. Hardware cost of a robot is very similar to a car, you know, but when it comes to R&D cost, you know, AI-related R&D cost will be way higher than that. of a vehicle, especially in the past. Cars used to partner with tier one suppliers when they can actually do the rule-based system and they can just borrow other people's already very established software and hardware solution, but that was in the past. In the AI era day and age, you cannot do it the way that it was for especially human robots development. And when it comes to operational costs, it's a completely different world as well, because there's different data for different scenarios, and when you apply it and deploy it in a new scenario, you need to retrain the AI and the robots as well. And there are two types of humanoid robots. You have the general purpose, you have the specialized purpose robots as well. What we mean by general purpose, basically, if you've attended college, you know that you first will take some general purpose courses that allow you to have some transferable skills and general knowledge. However, if you want to specialize in investment, for example, or in medical practice, that will require a lot of special training, right? And it's the same for humanoid robots. So I believe that when it comes to mass production, there will be chances or, you know, mass production will allow us to continue to drive down the bomb costs for humanoid robots. However, software and operational costs will continue to improve depending on what model and also what application scenarios that we're talking about. And we are different from a lot of the other companies out there because we do not target academia. We actually target actual deployment and commercialization. And we will first go into the commercial scenario, you know, before going to industrial and lastly household. Commercial application actually requires a lot of motion control, especially for the full body control, which is something that we're really good at and leading the industry. Industrial application will focus more on the hand movement and the agility of both hands, which we are working towards. And then household will be the most challenging scenario to actually deploy human robots in. And so we'll gradually move towards commercialization from commercial to industrial and to household. Thank you.

Thank you.

Your next question comes from Tina Ho from Goldman Sachs.

Please go ahead.

My first question is about our overseas business. 我想要问一下就是在今年我们这个销量翻倍 我们主要靠哪些市场来实现 然后包括明年后年就27、28年的时候 我们在靠一些什么样的一些产品或者是技术 去进一步去突破海外市场 还有一个小问题是咱们的这个昆蓬 超级增程的这些车型什么时候开始出海 So thank you for taking my question, and congrats on a strong set of results. So my first question is regarding our overseas expansion, wondering which will be the key markets for us to achieve the 100% volume growth in overseas this year, and also looking at 2027 and 2028. What kind of products or technology are we going to introduce in order to drive further expansion overseas? And then also related to that, When will we start to launch our Kunpeng super hybrid platform models in the overseas market?

Thank you. Brian, let me address your question.

In terms of a key market, obviously we start our global journey in Europe, and now the pan-European market, I would say, represents our largest market. regional market. Approximately 50% of our overseas volume is in Europe. And we actually saw our premium position EVs in a number of countries, especially in northern countries have a leading position in their categories. And also we saw very encouraging growth in large markets like Germany, France and Great Britain. So I think obviously Europe will represent a key focus for us. And then after Europe, Southeast Asia last year also represented a large, important growth market for us. We saw very, very exciting growth for Thailand, for Indonesia, for Malaysia, and we continue to see that market become a very key growth point for our business. Outside of these two large markets, additional emerging markets like Middle East, like Central Asia, like Latin America also has great growth potential and we actually have established operations last year and we see continued penetration will lead to also good growth prospects in those emerging markets. So how do we really accelerate that development? I think there's actually a number of factors contributing to our growth prospects. First of all, as you mentioned, is actually we are launching more products that covers wider market segments for these international markets. As we mentioned earlier, this year we intend to launch four new products. All four of them actually are global products. All four of them actually complements our current market segments, either on the lower end or the high end, and also has formats that's actually, I would say, very, very attractive and potentially useful for international markets as well. Also, as you just alluded to, this year we also intend to launch our extended range product line by the end of this year to select markets. I think that will also help us cover some of the international countries where charging facility as well as driving distances has been building enough, I would say, block for EV penetrations. So I think that's one of the areas that we're focused on supporting the global growth. On top of it, we also have been stepping up our localization efforts. For example, last year in our key markets in Europe, in Southeast Asia, we have started local production for some of our products. And I think that will definitely be a very important part of our growth strategy by providing opportunities more attractive products, better service closer to the customers with local production facilities and supply chain associated with that. We also intend to launch our smart driving technologies along with, obviously, the expected DCAS regulation in place by the end of the year. Hopefully, by early next year, our autonomous driving BLE 2.0 system can be utilized in international markets such as Europe and select markets in Southeast Asia as well. And on top of it, I think that there's other things we have to do as a company in organizational structure development, talent recruitment, brand building and marketing, as well as we actually are starting to roll out our proprietary charging facilities in select countries as well. All those efforts are intended to increase our competitiveness in international markets and elevate, expand brand awareness and positioning. So we have a very exciting prospect in the next, I would say, three to four years. We anticipate the global overseas market will represent much faster market growth compared to our overall growth rates, as well as becoming a core profit center for our business.

Okay, thank you. I look forward to seeing you in the overseas market. uh uh My second question is regarding our AI investment and then overall computing power. So as Mr. He Xiaopeng mentioned that this year, the physical AI investment will increase to 7 billion RMB. So wondering what is our compute power right now and also the plan for the next few years and whether these investments are all recorded in RMB or some of it will be also allocated to capital expenses.

Thank you. Because of this question,

I think the growth rate will be smaller in the next two years than it was in the last two years. Our investment in the AI field includes the current self-sufficiency, including our chip, RoboTaxi, and NGC, and so on. The only thing I think is that there may be more investment in NGC in the future, if NGC is further expanded after commercial mass production. All of our investments are once-in-a-lifetime as a development, not as a self-sufficiency. Thank you.

Thank you for your question. Since this involves long-term investment, I can only provide a relatively general response to your question. In the coming years, I believe that R&D in the automotive sector of our business will continue to converge and we will gradually stabilize that. But investment in AI will gradually increase and will trend upwards because of our conviction in AI development. However, the R&D spending will become more efficient. So in the subsequent years, the growth rate of our AI R&D investment will not be as significant or aggressive as previous years. Also, our major investment will include autonomous driving smart carpets and also RoboTaxi and humanoid robots among multiple categories. and capabilities. And so in the future, after mass production of humanoid robot begin to launch, you know, we also see greater investment in humanoid robots as well. And all of those investments will be booked as R&D expenses and not capex.

Sorry, can you go on?

Sorry. I think no one can give a clear and long-term judgment on the scale of computing today. We are now at the scale of tens of thousands of cards. I believe we must reach the scale of 100,000 cards. In the world of physical AI, I think the number of data in the world of physical AI is about 1,000 times that of digital AI, and the number of robots is about 1,000 times that of cars. We will use new methods in the future to solve the problem of computing AI.

我认为算力是一个长期,算力跟电力都是一个长期要解决的一个大问题,但是我觉得我们有方法,在以后会跟大家分享,谢谢。 Regarding the second part of your question, I think nobody nowadays can pinpoint exactly, you know, what is the compute power, you know, kind of requirement in the long term for physical AI. And right now we have thousands of GPU in the future. I think we need to at least reach 100,000 units for computing power accumulation for physical AI data training practices, at least that number. And I believe that cost training computing power requirement is going to be a thousand times for digital, you know, over digital AI and humanoid robot or robotics training will be a hundred, sorry, 1000 times of that of vehicles computing power requirements or standards. And in the future, definitely, I believe new methodology will be used to solve the computing power issue or shortage for physical AI and also the infrastructure, including power generation as well. Those will be long-term challenge for physical AI. However, we believe that we're going to figure something out to address that in the long term.

Thank you.

Thank you. Your next question comes from Ming Lee from Bank of America. Please go ahead.

Hello, Mr. Xiaopeng. I'm Ming from Meining Securities. I have two questions for you. The first question is about the current development and testing progress of Robotaxi. 然後今年有哪些重要的里程碑節點 包括這個有安全員載客 無安全員載客的運營目標 預計在什麼時候跑動 我會用英文翻譯一下 So what is our current progress of robot taxi testing? What do you expect the important milestone and the timing for our robot taxi business with safety driver and without safety driver in operations?

Yes, thank you. I have already answered it. We believe that in about one to three years, we will enter the complete automatic driving capability of the software layer, including China and the whole world. The progress of our hardware related to RoboTaxi is now very smooth. We are trying to solve the policy problem, because the policy needs to be carried out step by step, from the test license, from the process of safety to non-security. We also need to robot taxi的一些全球化的运营上去做一些额外的少量的研发我相信今年的下半年我们会展开有安全员的在客我们在明年年初我们希望能够无安全员并且在更后面我们会开放我们的整个的体系让更多的来自于中国来自全球的合作伙伴用我们的平台用我们的技术和产品能够去提供各个地方不同的 Thank you.

Thank you. I think I've already touched upon that question. We believe that a full, fully autonomous driving capability on the software side will actually arrive in about one to three years time, not only in China, but also around the world. Our hardware development regarding that has been moving and progressing smoothly. And right now, what is left is to address the regulatory requirements. And I think it will take time to do it step by step, you know, to obtain the testing license, and also to move from having a safety driver on board to going without a safety driver on board. And we are also doing some extra R&D for RoboTaxi global operation as well. And so I believe that by second half of this year, we should be able to do some test drives with a safety driver on board. And by the beginning of next year, hopefully we can do without the safety safety driver on board. And in the future, we're going to open the whole system, you know, to cover not only China, but also around the world, partnering with reliable operational partners around the world using our platform, our technology, and also our products so that we can provide this global taxi capability or autonomous driving capability for users around the world. Thank you.

What do you expect the scale or coffee number of our Robotaxi business in the future? And which models will provide the Robotaxi service?

I think the expansion of Lobotaxi in the future, I think the most important thing is based on the pace of society and the permissiveness of policies. I think in the next two years to maybe three or four years, it will be a priority in some countries and regions, and then grow quickly. I think what kind of model will start Lobotaxi in the future? I think this question, maybe in two or three years, we will think about it. Regarding this question, I think the

expansion or the actual deployment of RoboTaxi services depend on the regulatory development. And I think in the future, all parts of this ecosystem need to quickly develop. And what kind of vehicles or what kind of model will support RoboTaxi? I think in the couple of years, there will be a lot of thinking and exploration. And I think fundamentally, we need to really figure out whether or not RoboTaxi should be driven by cars or should be driven by robots. And I think in some time from now, people will actually see clearer, with very clear results or answer to that.

Thank you.

Thank you. As there are no further questions now, I would like to turn the call back over to the company for any closing remarks.

Thank you once again for joining us today. If you have further questions, please feel free to contact Expo's investor relations through the contact information provided on our website for the peace and financial communications.

This concludes today's conference call. You may now disconnect your line. Thank you.