Robots are being widely deployed—have automakers unlocked a second growth curve?
XPeng Group Q2 2026 Earnings Live Stream
[AI Key Takeaways]
Financial Performance
- Total revenue for the second quarter reached RMB 19.74 billion, up 8% year-over-year and 51.5% quarter-over-quarter
- Automotive sales revenue amounted to RMB 17.05 billion, representing a 1% year-over-year increase and a 55% quarter-over-quarter rise
- Revenue from services and other segments totaled RMB 2.7 billion, surging 93.9% year-over-year and 32.6% quarter-over-quarter
- Gross margin stood at 20.7%, compared to 17.3% in the same period last year and 20.6% in the previous quarter
Business Progress
- XPeng's robotics business completed its first round of financing, raising over USD 900 million and achieving a post-money valuation exceeding USD 6.3 billion
- Vehicle deliveries in the second quarter reached 103,295 units, a 65% increase from the previous quarter
- Overseas sales surpassed 20,000 units for the first time, marking an 81% year-over-year growth, while overseas revenue contribution exceeded 25%
- Robotaxis equipped with VLA technology have completed over 2,000 internal test orders in Guangzhou
Next Quarter Guidance
- Q3 deliveries are projected to range from 115,000 to 121,000 units, representing a quarter-on-quarter growth of 11.3% to 17.1%
- Q3 revenue is expected to be between RMB 21.7 billion and RMB 23.4 billion, marking a quarter-on-quarter increase of 9.9% to 18.5%
- Monthly sales are projected to exceed 60,000 units in the fourth quarter
- The robotics business plans to enter mass production by the end of 2026, with an official market launch in 2027
opportunity
- Product Innovation: The advanced general-purpose humanoid robot, Ion, features 76 degrees of freedom and is powered by three Turing AI chips
- Market Expansion: Overseas operations are growing rapidly, with an average selling price exceeding EUR 40,000
- Strategic Partnerships: Collaborating with top-tier investment firms such as IDG Capital, Hillhouse Ventures, Tencent, and Alibaba
- Operational Efficiency: Over 85% of supply chain partners overlap with the automotive business, realizing synergies
Risks
- Operational Disruptions: Extreme weather and supply chain issues have impacted the pace of production ramp-up, affecting delivery speeds
- Market Competition: The barrier to entry in robotics technology is extremely high, requiring unique capabilities across multiple dimensions.
- Economic Volatility: Facing industry-wide pressure from rising overall costs.
[AI Conference Transcript]
Operator
Hello ladies and gentlemen. Thank you for standing by for the Second Quarter 2026 Earnings Conference Call for X Pun Inc. At this time, all participants are in 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 C, Head of Capital Markets of the company. Please go ahead, Alex.
Alex C
Thank you. Hello, everyone and welcome to Expo second quarter 2026 earnings conference call. Our financial and operating results were issued by our news file services early today and available online. You can also view the earnings press release by visiting the IR section of our website at ir.shahom.com.
Participants of today's call from our management team will include Co founder, Chairman and CEO, Mr. Hershel Peng, live Chairman and President, Doctor Brian Group, Vice President of Finance and Accounting, Mr. James Wu and myself. Management will begin with prepared remarks and the call will conclude with a QA session. A webcast replay of this conference call will be available on the IRS section of our website.
Before we continue, note that today's discussion will contain forward-looking statements made under the Safe Harbor provisions of the US Private Securities Litigation Reform Act of 1995. Forward-looking statements involve inherent risk and uncertainties. As such, the company's results may be materially different from the views expected today.
Further information regarding these and other risks and uncertainties is included in the relevant public filings of the Company as filed with the US Securities and Exchange Commission. The Company does not assume any obligation to update any forward-looking statements except as required under applicable law.
Also note that Expo's earnings press release and this conference call includes the social unauthited GAAP financial measures as well as unauthited non GAAP financial measures. Next post earnings press 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. Hershoppang.
Hershel Peng
Hello everyone. I am pleased to share with our shareholders and investors that we have just announced the first round of financing for XPeng's humanoid robotics business. The robotics unit has raised over $900 million at a post-money valuation exceeding $6.3 billion, setting a new record for private financing in China's humanoid robotics industry.
This round was initiated by top-tier global investment institutions, led by IDG Capital with participation from Gao Rong Capital. We also received strong support from strategic investors Tencent and Alibaba. The size and valuation of this initial financing round have broken industry records, fully demonstrating the capital markets' high recognition of XPeng's leading position in physical AI, its technological roadmap, mass production capabilities, and long-term commercial value.
I believe this financing will not only provide ample capital support for the R&D and mass production of our advanced general-purpose humanoid robots but also accelerate our progress toward the 'ChatGPT moment' for physical AI. It will also bring more strategic resources to expand the robotics ecosystem and application scenarios. As a global leader in physical AI, I believe XPeng will not only lead the large-scale deployment and commercialization of autonomous driving worldwide but also build the most valuable embodied AI company. I am delighted that today we have taken another important step forward.
Good evening everyone. I am pleased to share with our shareholders and investors that we have just announced the first round of financing for XPeng's robotics business. The business raised over $900 million at a post-money valuation of over $6.2 billion.
This round was initiated by leading global investors, led by IDG Capital with participation from Gao Rong Ventures, and supported by Tencent and Alibaba as strategic investors. Both the size and valuation of the first round of financing have set a new private financing record in China's humanoid robotics industry, underscoring the capital markets' strong endorsement of XPeng's leadership in physical AI technology, its roadmap, ability to manufacture at scale, and long-term commercial value.
The financing will provide ample capital to support the mass production and continued development of our advanced general-purpose humanoid robot, IRON. We will accelerate our progress towards the ChatGPT moment for physical AI, while bringing additional strategic resources to strengthen the robotics ecosystem and expand real-world applications.
As a global leader in physical AI, XPeng will not only lead the large-scale deployment and commercialization of autonomous driving worldwide but also build the world's most valuable humanoid robotics company. Today, we are very happy to see that we have taken another important step towards that goal.
Since June this year, in addition to serving as CEO of XPeng Motors, I have personally taken on the role of CEO for the robotics business. Over the past 12 years, XPeng has consistently adhered to full-stack software and hardware capabilities, building a solid technological and commercial foundation for the era of physical AI. We are now able to mobilize the Group's advantageous resources for collaborative efforts, integrating capabilities accumulated in our automotive business—such as supply chain management, global layout of automotive-grade manufacturing, and underlying capabilities from our autonomous driving business, including Turing AI chips, algorithmic data, and kinematic models for the physical world—and applying them to the robotics segment. I believe this will definitely help XPeng Robotics achieve mass production and commercialization more quickly.
Since June, alongside my responsibilities as CEO of XPeng, I have also taken on the role of CEO of our robotics business. Over the past 12 years, XPeng has remained committed to full-stack in-house R&D across both software and hardware, building a solid technological and commercial foundation for the physical AI era.
We are able to consolidate and integrate the strengths and resources of the entire group. These include the supply chain, automotive-grade manufacturing capabilities, and global footprint developed through our automobile business, as well as the Turing AI chips, AI infrastructure, and world-class foundation models developed through our other businesses. By applying these capabilities to our robotics business, I believe we can accelerate the mass production and commercialization of XPeng's humanoid robots.
XPeng has been engaged in robotics R&D for eight years. Over time, I have come to realize that the technological challenges and innovation difficulties associated with advanced general-purpose humanoid robots are far greater than those for smart electric vehicles—I estimate the gap is at least twentyfold. To achieve such significant innovation, we must conduct comprehensive, in-house R&D and system integration with both breadth and depth across design and styling, hardware and chips, software and AI, data and management, as well as quality and manufacturing. I believe this is the only way to successfully develop robots.
I believe XPeng Robotics may be the only company among all Chinese robotics firms to conduct in-house R&D across its entire system. Because of this, I believe XPeng's 'Iron' robot differs fundamentally from current competitors on the market, offering unique capabilities and overall balance in multiple areas.
We have been conducting research and development in the field of robotics for more than eight years, and I have always believed that the technological challenges and level of innovation required for advanced general-purpose humanoid robots are far greater than those for smart EVs—by at least 20 times.
To accomplish something this difficult, we need broad and deep R&D and system integration capabilities spanning design and styling, hardware and chips, software and AI, data and control systems, as well as quality in manufacturing. Only then can we succeed.
XPeng is the only robot company in China with such comprehensive in-house R&D capabilities across the entire technology stack. This is why XPeng's Iron stands apart fundamentally from other humanoid robots currently on the market, with differentiated capabilities across multiple areas.
Currently, our full-stack in-house technology covers the robot's body, brain, cerebellum, data, and related infrastructure. In terms of the body, XPeng's Iron features the industry's most leading theoretical status and design. We pioneered the industry's first fully enclosed flexible lattice structure, combining aesthetics with safety. The whole body has 76 degrees of freedom, with 21 degrees of freedom in each dexterous hand, both of which are at the industry's leading level.
XPeng Motors has independently designed and developed an AI-native hardware platform and virtually all core components around embodied intelligence, including chips, controllers, key motion modules, and dexterous hands. Furthermore, leveraging our established R&D and manufacturing system for smart electric vehicles, we have built the capability to deliver automotive-grade robot quality and achieve large-scale mass production and delivery.
Currently, our full in-house technology stack covers Iron's body, brain, cerebellum, data, and infrastructure. On the hardware front, XPeng Iron features the industry's most human-like form and design. XPeng pioneered the industry's first fully enclosed flexible lattice structure for Iron, combining aesthetic appeal with enhanced safety, with 76 degrees of freedom across the body.
And the 21 degrees of freedom in each hand are both at the industry's leading level. XPeng has independently designed and developed an AI-native hardware platform and all core components specifically for embodied intelligence, including chips, controllers, motion modules, and dexterous hands. Leveraging our established smart EV R&D and manufacturing capabilities, we can achieve automotive-grade quality and build the capability to manufacture and deliver at scale.
In terms of intelligence, XPeng's Iron is equipped with three Turing AI chips, delivering an effective computing power of up to 2,250 TOPS. Leveraging what I believe is the highest edge computing power in the robotics industry, XPeng's Physical AI foundation model has been successfully deployed on-device. This enables the robot to autonomously complete complex tasks without remote operation, while ensuring low-latency inference and high data security locally.
The highly anthropomorphic hardware platform of the robot naturally possesses stronger advantages in data scaling. It can maximize the reuse of behavioral data generated from human daily life and rapidly adapt to a wide range of application scenarios and tools designed for humans. I believe that as XPeng's Allen enters mass production first, we will also gain access to massive amounts of data from value-added scenarios and human demonstration data.
All of this will accelerate the training and iteration of our AI models, continuously enhancing our model capabilities. This, in turn, will support the deployment of more robots into more scenarios, generating even more data. This creates a flywheel effect across data, models, and applications, accelerating the evolution of XPeng's Allen in the real world.
In terms of intelligence, XPeng Iron is powered by three Turing AI chips, delivering effective computing power of up to 2,250 TOPS, representing the industry's highest level of edge computing capability. XPeng's Physical AI foundation model runs directly on Iron, enabling it to autonomously perform complex work tasks without teleoperation, while ensuring low-latency inference and data security.
Furthermore, the highly anthropomorphic hardware platform offers natural advantages in data scaling. It allows us to maximize the reuse of behavioral data generated from people's everyday lives and rapidly adapt to a broad range of environments and tools designed for humans.
As XPeng Allen moves toward mass production and real-world deployment, we will gain access to vast amounts of real-world and human demonstration data, accelerating the training and iteration of our AI models. In turn, continuous improvements in model capabilities will allow Allen to enter more scenarios and generate more high-quality data, creating a flywheel effect across data, models, and applications that will accelerate Allen's evolution in the real world.
We expect XPeng's Allen to combine an ultimate theoretical design, the smartest AI brain, and the highest standards of safety quality. Only in this way can Allen become a trustworthy partner for humans, integrating into public work and daily life. Our mass-production version has recently achieved several important milestones. Starting next month, September, I will gradually showcase and introduce Allen's series of unique capabilities.
We plan to enter the stage of large-scale mass production by the end of this year, beginning with commercial scenario applications in XPeng's stores and parks. XPeng Allen will be officially launched in 2027, targeting external customers in retail and service sectors, with large-scale deliveries starting in China and overseas. Next year, monthly production capacity will rapidly increase to thousands of units based on market demand.
I believe the technical barriers for advanced general-purpose humanoid robots are extremely high, and supply is quite scarce. Therefore, the revenue and gross profit contribution of each Allen unit over its entire lifecycle—including hardware sales and software-related AI model services—will generate recurring income. As a result, these figures will certainly be significantly higher than the current average selling price and gross margin per vehicle in our automotive business.
I look forward to the rapid commercialization of humanoid robots domestically and overseas after mass production, bringing us substantial gross profit growth. This will support greater investment in our Physical AI, allowing us to continuously expand our technological lead.
XPeng Iron combines an exceptionally human-like design, the most intelligent AI brain, and the highest standards of safety and quality. Only by doing so can Iron become a trusted companion to people and truly integrate into everyday working life. We have recently achieved several major milestones in the development of the mass-production version.
Starting in September, we will unveil and demonstrate a series of distinctive capabilities. We plan to enter scaled production by year-end, with initial commercial deployment in XPeng stores and campuses. In 2027, the XPeng Era will officially launch, beginning large-scale deliveries to external customers in the retail and service sectors both in China and overseas.
Next year, monthly production capacity can rapidly ramp up to several thousand units in response to market demand. I believe the technological barriers to advancing general-purpose humanoid robots are exceptionally high, while the supply of high-quality humanoid robots remains limited.
As a result, the lifetime revenue and gross profit contribution of each unit, including hardware sales and recurring revenue from upgrades to its AI model capabilities, will be substantially higher than the current average selling price and gross profit per vehicle in our automotive business. I expect the commercialization of humanoid robots to scale rapidly in China and overseas following mass production, generating meaningful gross profit growth, supporting our investment in physical AI R&D, and further widening our technological lead.
Regarding our automotive business, in the second quarter, XPeng Motors delivered 103,295 vehicles, representing a 65% quarter-over-quarter increase, and achieved year-over-year growth ahead of the industry's recovery. Amidst industry-wide pressure from rising costs, our operations remained robust. Thanks to breakthroughs in premiumization and international expansion, XPeng's gross margin in the second quarter remained above 20%.
Among them, the tech-luxury flagship XPeng G9 stood out among the series of large six-seat SUVs launched this year. Domestic deliveries exceeded 7,000 units in July, making it one of the top three best-selling new energy SUVs priced above RMB 300,000. Furthermore, the first SUV in the Mona series, the L03, became a phenomenal hit upon launch, with orders breaking historical records for all XPeng models.
Moreover, even in July and August, which have already passed in the third quarter, we saw XPeng Motors' new order volume increase by over 50% quarter-over-quarter, hitting a new historical high. I believe that due to extreme weather and supply chain impacts, our current production ramp-up pace has indeed affected delivery speed. Here, I would like to express my sincere gratitude to all car owners for their patient waiting.
We have initiated double-shift production for the Mona L03 and are going all out with our supply chain partners to accelerate capacity expansion. I am confident that deliveries of the L03 will increase significantly in the coming months, continuing an upward trend.
Now I would like to return to our core automotive business. In the second quarter, our vehicle deliveries reached 103,295 units, up 65% quarter-over-quarter, and we achieved year-over-year growth ahead of the broader industry. Despite industry-wide cost pressures, our operations remained resilient, supported by our progress in the premium segment. The company's gross margin remained above 20% in the second quarter.
Our tech-defined luxury flagship model, the G9, stood out among the wave of large six-seat SUVs launched this year. Domestic deliveries exceeded 7,000 units in July, making it one of the top three models in China's NEV SUV segment priced above RMB 300,000. The Mona L03, the first sedan in the Mona series, became a breakout hit immediately after its launch, with orders setting a new record for any XPeng model.
New uncancelable orders increased by 50% quarter-over-quarter to a record high. Extreme weather and supply chain disruptions affected our production ramp-up and delivery pace. Here, I would like to express my sincere appreciation to our customers for their patience. We have started shifting production to the Mona L03 and are working closely with our supply chain partners to accelerate capacity expansion. I expect L03 deliveries to increase substantially over the coming months and continue an upward trend.
The consecutive success of the G9 and L03 gives us greater confidence in our future products. We will translate our leading advantages in intelligence, aesthetics, and quality into higher sales targets and stronger brand momentum. Our flagship five-seater SUV, the G9L, will officially launch and begin deliveries in September. Meanwhile, XPeng's L05 will also launch in the domestic market in the fourth quarter. I believe that with the launch of these four new SUV models, we will achieve better coverage of the main SUV segments. I expect XPeng's sales to rise significantly in the fourth quarter, with monthly sales exceeding 60,000 units.
The success of the G9 and the Mona L03 gives us more confidence in our upcoming models. We are translating our competitive strengths in best-in-class intelligence and standout design into higher sales targets and stronger brand momentum. Our flagship 5-seat SUV, the G9L, will officially launch and begin delivery in September. The Mona L05 will also launch in China in the fourth quarter, alongside the launch of four brand-new SUV models.
We will cover all major SUV segments. We believe deliveries will increase significantly in the fourth quarter, with monthly deliveries targeting more than 60,000 units.
Global business is the second growth curve for XPeng Motors and an important driver for improving profitability. In the second quarter, our overseas sales exceeded 20,000 units for the first time, an 81% year-over-year increase. In the first half of this year, XPeng's overseas business accounted for more than 25% of total revenue and achieved high operational quality. With an average selling price exceeding €40,000, both per-vehicle revenue and gross profit are at the forefront of Chinese automakers going global.
In July, the XPeng Mona M03 garnered significant attention following its global launch in Munich, Germany. Its intelligent features, styling, and space received high praise from overseas consumers. I believe the L03 will become XPeng's first major model to lead sales in multiple global markets. We expect overseas deliveries to begin in the fourth quarter, driving the company's quarterly overseas sales to exceed 40,000 units. In 2027, we will also launch several star models overseas, including extended-range electric vehicles, to further expand XPeng's geographic coverage and market share.
Our international business is the second growth engine for XPeng Motors and an important driver for improving profitability. Overseas quarterly deliveries exceeded 20,000 units for the first time in the second quarter, up 81% year-over-year. In the first half of the year, our international business accounted for more than 25% of total revenues.
Furthermore, our overseas operations boast exceptional quality, with the average selling price of our exports exceeding €40,000, placing our value, revenue, and gross profits at the forefront of Chinese automakers. Since its global launch in Munich in July, the Mona L03 has attracted significant attention and earned high praise from overseas consumers for its intelligent technology, distinctive styling, and spacious interior.
I believe the Mona L03 will become XPeng's first major model to achieve leading sales across multiple international markets. Overseas deliveries of the Mona L03 are expected to begin in the fourth quarter, driving company-wide quarterly overseas deliveries to exceed 40,000 units. By 2027, we will also introduce multiple star models, including extended-range EV models, in overseas markets, further expanding our geographic coverage and market share.
Starting from the end of August, our second-generation VLA large model will undergo its first major upgrade. In the field of autonomous driving, we have once again verified that XNGP brings significant improvements to experience, safety, and generalization. In this new version 6.3, our model parameters have expanded by 3.5 times, creating an order-of-magnitude gap compared to the small models commonly used in the industry.
The new version also enhances our perception capabilities, with sensitivity improved by 300%, and introduces ultra-long-horizon reasoning and predictive abilities. The operating frame rate is now industry-leading, enabling the AI driver to see accurately, think critically, and respond quickly. Furthermore, this new version achieves the first-ever integration of VLA (Vision-Language-Action) and VLM (Vision-Language Model) for unified smart cockpit intelligence. It brings down certain L4-level experiences from XPeng's Robotaxi to consumer vehicles; for instance, users can pull over to the curb via voice commands. I believe this second-generation VA system can autonomously and safely locate roadside parking spots and complete parking maneuvers while maintaining autonomous driving mode.
Starting from the end of August, we will roll out a major upgrade to VLA 2.0, once again validating Stating Law in others and delivering substantial improvements in both user experience and safety. With the new major version 6.3, the number of parameters in the on-device VLA 2.0 model will increase by 3.5 times, placing its parameter count an order of magnitude above that of small models commonly used in the industry.
The new version delivers a 300% improvement in perception sensitivity, introduces ultra-long-horizon reasoning and predictive capabilities, and operates at an industry-leading frame rate. This enables the AI driver to see accurately, think ahead, and respond quickly. The new version also integrates smart cockpit capabilities powered by VLA and VLM, bringing select Level 4 capabilities developed for the XPeng Robotaxi to our passenger vehicles.
One example is voice-activated pullover parking. Users simply give a voice command, and VLA 2.0 will autonomously find a suitable roadside parking space and park the vehicle without exiting autonomous driving mode.
Recently, my colleagues and I test-drove XPeng's second-generation VA and the latest versions from global competitors in China, the US, Australia, and Europe. It is evident that on mainland roads, XPeng's second-generation VA is already on par with competitors' products. However, in complex scenarios such as narrow roads, interactive negotiations, and internal campus parking lots, the user experience of XPeng's new VA 6.3 version is significantly better.
We support direct navigation from route guidance to parking spots, ensuring both efficiency and safety. I believe that with upgrades in computing power and model capabilities, several major versions following this generation of VA will unleash more powerful comprehensive capabilities. We will achieve an L4-level autonomous driving experience in mass-produced vehicles, rapidly surpassing competitors and establishing a generational lead in intelligent technology.
Recently, together with my colleagues, we test-drove XPeng's VLA 2.0 and the latest offerings from a leading global peer in China, Europe, and North America. In my view, VLA 2.0 is already on par with the world's leading competitors on major roads. In challenging scenarios such as narrow roads, interactive negotiations, and campus parking facilities, the user experience delivered by VLA 2.0 is even superior.
It can navigate directly to a parking space with both efficiency and safety. I believe that as we upgrade computing power and model capabilities, VLA 2.0 will develop even more powerful features over the next several version updates, delivering an L4-level autonomous driving experience in mass-produced vehicles, surpassing peers and establishing a generational lead.
Meanwhile, we are going all out to accelerate the global deployment of the second-generation VA. Recently, my team and I completed local acceptance testing for VA in Germany. Honestly, I was very excited because our model achieved a user experience in Germany that closely matched that in China, with almost no increase in localized training data. I believe this demonstrates a clear lead over competitors in terms of global multi-region generalization capabilities.
Our goal is to secure regulatory approval in Europe by the first half of 2027 and simultaneously begin delivering the second-generation VA to overseas users in more countries. We hope to allow global users to enjoy a safe, comfortable, and convenient intelligent mobility experience. Once the second-generation VA is deployed overseas, we will be delighted to compete directly with the most advanced assisted driving systems from other global peers. Furthermore, XPeng will actively explore new software-based business models, creating a positive cycle where commercialization and R&D mutually reinforce each other.
We continue to enhance certain model capabilities of VLA 2.0 while accelerating its global deployment. Recently, my team and I completed on-road validation of VLA 2.0 in Germany. Particularly encouraging was that the model, primarily trained on data from China, performed nearly as well on European urban roads as it did in China, with almost no additional local training data.
We aim to secure regulatory approval for VLA 2.0 in Europe in the first half of next year and roll out VLA 2.0 to deliver a safer, more comfortable, and more convenient driving experience to users worldwide. After deploying in overseas markets, VLA 2.0 will compete directly with other world-leading solutions and become a key differentiator, defining the competitive advantage of our global products. At the same time, we will actively explore new software-based business models, creating a virtuous cycle where commercialization and technological development reinforce each other.
As of now, pre-installed mass-produced vehicles equipped with VLA have completed over 2,000 internal test orders in Guangzhou. We have also validated the full end-to-end process for supervised passenger demonstration operations, laying the groundwork for commercial operations. We have recently completed the development of our cloud-based remote takeover platform, with the goal of achieving driverless passenger operations without safety operators next year.
In 2027, XPeng Motors will continue to enhance its technological and cost competitiveness. By partnering with leading mobility platforms domestically and internationally, we will rapidly expand our business scale in core cities in China and overseas. We aim to create greater commercial value through vehicle sales, technology services, and revenue sharing from operations.
As of now, our pre-installed mass-produced Robotaxi powered by VLA 2.0 has completed more than 2,000 internal test orders in Guangzhou and validated the full end-to-end process for trial passenger operations, laying the groundwork for commercial operations. Recently, we have completed the development of our cloud-based remote takeover platform, and our goal is to begin driverless operations without a safety operator in the car next year.
In 2027, XPeng Motors will continue strengthening the technology and cost competitiveness of our Robotaxi while partnering with leading domestic and international mobility platforms to expand our Robotaxi business across key cities in China and around the world. This will create greater commercial value through vehicle sales, technological services, and revenue sharing from operations.
At the same time, we believe that the advancement of XPeng's Physical AI technology requires not only technological breakthroughs but also the creation of an open, collaborative ecosystem that ensures mutual benefits for all stakeholders. To accelerate the commercialization of Physical AI, we have recently established a group-level Business Development team. We are currently actively negotiating with partners both overseas and domestically, aiming to bring our industry-leading Turing AI chips, VLA models, and even humanoid robot technologies to the global market more quickly.
In the meantime, we believe that the large-scale application of Physical AI requires more than just technological breakthroughs; an open and collaborative technology and business ecosystem that creates value for multiple participants is equally important. To accelerate the commercialization of Physical AI, we recently established a group-level business development team and are actively engaging with partners in China and overseas to bring our industry-leading Turing AI chips, VLA 2.0, Robotaxi, and humanoid robot technologies to global markets more quickly.
I expect XPeng's deliveries in the third quarter of 2026 to be approximately 115,000 to 121,000 units, representing a quarter-on-quarter growth of 11.3% to 17.1%. Revenue is projected to be around RMB 27.1 billion to RMB 28.4 billion, reflecting a quarter-on-quarter growth of 9.9% to 18.5%. I believe that starting from the second half of the year, XPeng will enter a phase of enhanced momentum driven by the resonance of multiple business lines. Not only will domestic and overseas vehicle sales reach new highs, but we also believe our autonomous driving technology will achieve broader leadership, with advanced intelligent driving assistance technologies being rolled out globally in batches.
I also look forward to us being the first globally to achieve mass production and commercialization of advanced general-purpose robots. Overall, by building new technological capabilities and business models around the AI ecosystem, we aim to create greater value for our customers and shareholders.
For the third quarter of 2026, we expect deliveries to be approximately 115,000 to 121,000 units, representing quarter-over-quarter growth of 11.3% to 17.1%. Revenue is expected to be approximately RMB 21.7 billion to RMB 23.4 billion, representing quarter-over-quarter growth of 9.9% to 18.5%.
I believe XPeng is entering a period of accelerating momentum across multiple business lines. In the second half of the year, both domestic and overseas vehicle deliveries are expected to reach new highs. We also expect to be among the first globally to achieve scaled mass production and commercialization of advanced general-purpose humanoid robots. Furthermore, we anticipate being among the first to deploy advanced AI technologies in overseas markets, establishing new business models around the physical AI ecosystem to create greater value for customers and shareholders worldwide.
Thank you all. Next, our VP of Finance, James, will introduce our financial results for the second quarter of 2026.
James Wu
Thank you, Xiaopeng. Now let me provide a brief overview of our financial results for the second quarter of 2026. Unless otherwise stated, I will reference figures in RMB only. Our total revenues were RMB 19.74 billion for the second quarter of 2026, an increase of 8% year-over-year and an increase of 51.5% quarter-over-quarter.
Revenues from vehicle sales were RMB 17.05 billion for the second quarter of 2026, an increase of 1% year-over-year and an increase of 55% quarter-over-quarter. The quarter-over-quarter increase was mainly attributable to higher vehicle deliveries. Revenues from services and others were RMB 2.7 billion for the second quarter of 2026, representing an increase of 93.9% year-over-year and an increase of 32.6% quarter-over-quarter.
The year-over-year and quarter-over-quarter increases were primarily attributable to: first, increased revenues from technical R&D services rendered to the Volkswagen Group due to the successful achievement of certain key milestones; and second, parts and accessory sales. Gross margin was 20.7% for the second quarter of 2026, compared with 17.3% for the same period of 2025 and 20.6% for the first quarter of 2026.
Vehicle margin was 12.1% for the second quarter of 2026, compared with 14.3% for the same period of 2025 and 12.1% for the first quarter of 2026. The year-over-year decrease was primarily due to the transition between vehicle generations. R&D expenses were RMB 2.91 billion for the second quarter of 2026, representing an increase of 32.1% year-over-year and an increase of 0.3% quarter-over-quarter.
The year-over-year increase was mainly due to higher expenses related to the development of new vehicle models and AI-related technologies, as the company expanded its product portfolio to support future growth. SG&A expenses were RMB 2.5 billion for the second quarter of 2026, representing an increase of 15.2% year-over-year and an increase of 32.5% quarter-over-quarter.
The year-over-year increase was primarily due to higher marketing and advertising expenses. The quarter-over-quarter increase was primarily due to higher commissions paid to franchise stores and higher marketing and advertising expenses. As a result of the foregoing, the loss from operations was RMB 1.14 billion for the second quarter of 2026, compared with RMB 0.93 billion in the same period last year and RMB 1.87 billion in the previous quarter.
Net loss was RMB 1.34 billion for the second quarter of 2026, compared with a net loss of RMB 0.48 billion in the same period last year and a net loss of RMB 1.7 billion in the previous quarter. As of June 30, 2026, our cash position was RMB 14.48 billion. To be mindful of the length of the earnings call, I encourage listeners to refer to our earnings press release for more details on our second quarter 2026 financial results.
This concludes our prepared remarks. We'll now open the call to questions. Operator, please go ahead.
Operator
Thank you. If you wish to ask a question, please press *1 on your telephone and wait for your name to be announced. If you wish to cancel your request, please press *2. If you are using a speakerphone, please pick up the handset to ask your question. For the benefit of all participants on today's call, if you wish to ask management a question 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 you may follow up with your next question. Your first question today comes from Tim Sowell with Morgan Stanley. Please go ahead.
Tim Sowell
Hello, management. This is Tim. Thank you for taking my question, and congratulations on completing the financing for the humanoid robot project at such an impressive valuation. I have two questions related to robotics. My first question is a follow-up: Mr. He mentioned in the briefing that the second-generation robot is scheduled to enter large-scale mass production by the end of this year. Could you please share the plan for ramping up production capacity? It was mentioned that volume would reach several thousand units next year. Additionally, what is the total delivery target for 2027? As we consider financing, we would like to understand the underlying assumptions. That is my first question.
So my first question is about volume and production targets. What is the projected production capacity for XPeng's humanoid robots upon entering commercial mass production by late 2026? And what is the target delivery volume for fiscal year 2027? That's my first question.
He Xiaopeng
Thank you for your question. This is He Xiaopeng. From our perspective, there is a significant difference between robot production capacity and automobile production capacity. The automotive supply chain is extensive in both breadth and depth. However, for XPeng's Iron humanoid robot, we have not only conducted full-stack self-research and development across multiple areas but have also deeply integrated in-house manufacturing. Therefore, I believe the primary challenge for robot mass production in the early stage will be quality, while in the later stage, it will be sales. Overall, the required production capacity will be significantly smaller compared to the automotive scenario.
Regarding mass production, we plan to begin large-scale mass production by the end of 2026. After commencing mass production, we will start with internal pilot programs, initially deploying XPeng's Iron humanoid robots in XPeng stores to establish commercial use cases. In the first half of 2027, we will begin deployment in our own commercial scenarios and selected pilot environments, while continuously accelerating external deliveries.
As product R&D quality and supply chain maturity improve, we will continuously accelerate our production ramp-up in the second half of the year. We will expand our monthly production capacity to several thousand units or even higher, depending on market demand.
As product R&D quality and supply-chain maturity continue to improve, we will steadily accelerate our production ramp-up throughout the year and in the second half. Based on market demand, we plan to scale up monthly capacity to several thousand units, or even higher.
Regarding mass production in 2026, we expect it to commence by the end of that year. We anticipate scaled mass production for the Iron robot products. Consequently, we believe commercialization will begin within our own retail stores. In 2027, commercialization will expand across various self-operated scenarios, accelerating external rollout and R&D use cases.
We expect to start around mid-second half of next year. By 2027, we will accelerate R&D development, with mass production volumes reaching several thousand units per month, further increasing pace.
Finally, I would like to add that since XPeng's Iron robot deliveries target retail and service sectors, with gradual rollouts both domestically and overseas, our delivery figures will be as tangible and verifiable as our automotive sales. Our capabilities in commercial scenarios, such as intelligent shopping assistance, differ significantly from other robotics companies. I believe the actual deployment of these capabilities will quickly demonstrate stronger generalization across various industry roles, leading to more comprehensive commercialization. Thank you.
To supplement, the Iron robot will primarily be deployed in retail and service sectors, both in China and abroad. Similar to our core automotive business, our robot delivery and sales data will be authentic and genuine, reflecting the true quality of our robots and services.
I believe that compared to peers in the market, whether in terms of shopping assistance or intelligence levels, we will be superior and stronger. Additionally, our solution will be adaptable to a wider and broader range of environments.
Thank you, Mr. He Xiaopeng, for the detailed sharing. My second question concerns the demand side, which you mentioned earlier alongside quality. Could you please provide an estimate of the unit cost for the current mass-production version of the robot? What pricing strategy is planned going forward, and what gross margin level is anticipated after mass production ramps up?
My second question relates to unit economics and margin profiles. What is the estimated unit cost for the mass-production variant of the Iron robot? What go-to-market pricing strategy does management intend to deploy? And what is the anticipated gross margin trajectory, particularly after full-scale ramp-up? That is my second question. Thank you.
Hershel Peng
Thank you. For the mass production of Iron, to ensure innovation integration, quality, and capacity, we have independently designed and developed the entire native hardware platform and software system. Over 85% of our supply chain partners overlap with XPeng's current automotive supply chain. Therefore, we are confident that the cost competitiveness of XPeng's robots will be leading in the industry.
Currently in China, robot product pricing typically differs from automobiles, often ranging from 2.5 to 3 times the Bill of Materials (BOM) cost. As Iron is positioned as a general-purpose robot, I believe it has significant pricing flexibility and scarce supply. Consequently, its hardware gross margin will likely be substantially higher than our current automotive business margins. Furthermore, beyond hardware sales, we will generate additional revenue from AI model capabilities and software, thereby enhancing overall revenue and profitability. Thank you.
Thank you for your question. Regarding our mass-produced robots, specifically the humanoid models, we are focusing on innovation, quality, and capacity. For both our hardware and software, we are conducting R&D entirely in-house. As for the supply chain for these components, over 85% of our supply chain partners are actually shared...
...with the existing supply chain partners for our other business lines. I believe this gives our Iron Robot a leading competitive edge in terms of cost. Currently, robot pricing in the market is generally about 2.5 to 3 times the bill of materials (BOM) cost.
Given that these are general-purpose robots and supply in the market is very limited, I believe our hardware gross margins will definitely be better than those of our existing alternative businesses. Meanwhile, we are not relying solely on hardware sales; we will also generate revenue from different models, software services, subscriptions, etc. We believe all these avenues will contribute profits to our business.
Yes, that concludes my response to your question. Thank you.
Operator
Thank you. Your next question comes from Ming Sun Lee with Bank of America. Please go ahead.
Ming Sun Lee
Hello, Mr. Xiaopeng. I have two questions related to robots. First, which parts of your robot's foundation model can be highly synergistic with autonomous driving? Which modules are shared, and which are developed relatively independently?
Hershel Peng
Many in the industry say that robots might just need one brain, or one large model. While this may be possible many years from now, I think it is completely unfeasible at present. I believe there are mainly three types of models in robotics: first, ultra-fast models, operating at hundreds of frames per second; second, medium-speed models, such as those for autonomous driving capabilities, operating at about 10 to 20 frames per second; and third, slow models, or 'thinking' models, operating at about one frame per second. Currently, the models in XPeng Motors vehicles, such as our VOA and our multimodal VOM, are largely similar to those integrated into the robots.
For example, our VOA is currently developing roaming capabilities for non-navigated roads, such as underground parking lots. This follows the same logic as the future indoor roaming functionality of our robots. At the same time, we see that some robot models, such as the thinking VOT model and the open platform series, will create value for our automotive systems. These will be integrated into the XPeng Motors vehicle ecosystem next year or thereafter. In fact, there are many other interesting models within the integration, such as whole-body motion control models and various safety system models, including those for data privacy, fall detection, and power failure safeguards.
These are some unique models, but they share commonalities overall. Therefore, within our system, we will jointly develop related models under the broader XPeng Motors ecosystem. This includes more foundational layers, such as world model simulations and generative models for world models. We are developing these collaboratively. Thank you.
Thank you for your question. To address this, many in our industry argue that robots generally only need one 'brain' or a single large model, which would be sufficient. While this may become feasible in the distant future, I do not believe it is viable at present. Regarding the differences among large models, there are various types.
For instance, we have super-fast models operating at 100 frames per second (fps) or even several hundred fps. There are medium-speed large models operating at 10 to 20 fps. Additionally, there are slower large models, which we refer to as 'thinking large models,' operating at 1 fps.
Regarding BLA and BLM, these are adopted in our alternative business lines. I believe there are similarities; for example, the BLA currently used in our alternative business. When it comes to navigation on unplanned roads, this is quite similar to the roaming or movement capabilities of robots.
On the other hand, I also believe that certain cognitive capabilities of robots, such as those on our open platform launching next year, can be applied to our automotive business. Thus, there are definitely synergistic commonalities. Meanwhile, robotics offers unique perspectives, such as different safety modes.
For instance, safety models cover areas like data privacy, prevention of mechanical folding failures, and power management safety to prevent battery depletion. As you can see, there are commonalities, and we are developing all these features under the unified XPeng system.
Even so, if we look at the deeper underlying systems, there are many areas that are quite distinct, such as the development of generative models, as well as mimicking and simulation models. These represent both the similarities and differences.
Thank you. Before moving to my second question, I would like to add that beyond the models themselves, the governance and organization of AI infrastructure data are also shared across these two business lines. OK, thank you.
Apologies, before you proceed to your second question, I would like to supplement my previous point. Apart from the models, whether regarding AI applications or the overall architectural structure, these resources are shared across the two different segments of our business.
Now, for my second question, which is still related to robotics: Compared to other robotics startups, what differentiated advantages does XPeng Motors have in areas such as data collection, training, and closed-loop iteration?
Hershel Peng
Yes, thank you. That is an excellent question. I believe data will be a critical core component of Physical AI in the future and a key source of differentiated value. Many argue that if robots have sufficient data, they will perform well. While this may be a necessary condition, it is certainly not a sufficient one. Today, there are many companies developing autonomous driving vehicles, but few excel at it. Therefore, I believe the most crucial factor is how effectively data is utilized, including its quality, diversity management, and training processes.
Regarding XPeng's robotics data and autonomous driving data, I believe a core and critical advantage is XPeng's ten years of R&D and data experience in autonomous driving. By the end of this month, our latest version will achieve industry-leading capabilities, which I consider to be the most important factor.
Furthermore, our robotics data management, training, and entire infrastructure are developed within the same ecosystem as XPeng's automotive business. The only differences lie in hardware aspects such as data collection equipment. Therefore, I am highly confident in our robust closed-loop capabilities spanning from data collection to application.
As XPeng Allen enters mass production, we will collect more real-world scenario data and human demonstration data required for user interaction. All of this constitutes high-quality data. In contrast, low-quality data often reduces overall precision. We aim to leverage high-quality data to create a flywheel effect for our data models and applications, accelerating Allen's evolution in the real world. Thank you for your questions.
Thank you. That is a great question. And I do believe that in terms of the physical AI and in the future and data is of course the key. Many people say that as long as you have enough data and that will help with the integration of the services, it is a necessary condition. However, it is not yet the full condition and what we would see is that for Expo and what we are good at is that we are have much better data and the training of the data, we also have a higher quality of the data.
For Expo. For instance, we have been in the area of autonomous driving for over 10 years within not only our R\u0026D for the past over decade of experience as well as the data that we have collected. We are absolutely leading in the industry among our peers in terms of the robots, it is the same for the robots and management.
The robots, data management, data training and data quality, all of these are being developed under the same ecosystem as our company. And I believe that apart from the hardware. So being different in in terms of the collection of the data and to all the way to the application of the data at our company for our two different lines of business would be the same for I run again once at this products that become a mass produced and launched into the market.
Not only that. So we will be further continuing to collect the real world data as well as the human demonstrated data and which are both sets of high quality data. And this will further drive the development in R\u0026D of our products. And this is different from the low quality data and which are not helpful at all. So I believe this answer by having all of these that would actually be able to create a flywheel of the high quality data and R\u0026D that we are conducting and to continue to contribute to the development of our products. That's all my answer for your question. Thank you.
Ming Sun Lee
Thank you, Xiao Peng. That concludes my questions.
Operator
The next question comes from Jeff Chung at Citi. Please go ahead.
Jeff Chung
Ah, thank you, Mr. Xiao Peng and the management team, for this opportunity to ask questions. My first question is regarding XPeng's robotics. Why did you choose scenarios such as sales and tour guides as the initial deployment use cases? Who are the target customers, and why would they purchase ION? Most importantly, are there subsequent plans to expand into industrial and household scenarios?
My first question is about the why did Expand robotics select the salesperson, the tour guide scenario as the initial real world deployment? Who are the target customers? Why would customers buy ION? And most importantly are their follow up plans to expand into industrial and the home use cases? Thank you.
Hershel Peng
Thank you. Actually, the commercial scenarios envisioned for XPeng's robots differ significantly from those of many other companies. While many assume entry via factories or households, primarily through a B2B approach, we believe our target customers are mass consumers (To-C) and small-to-medium businesses (SMBs). We intend to enter the market through commercial applications first. Naturally, following our commercial entry, we will expand into industrial and household sectors, utilizing a very limited number of SKUs to achieve this.
So, why start with commercial applications, specifically the 'three islands' scenario? I believe the most critical capability lies in how we categorize our robots' comprehensive abilities into four key areas. We believe that within both Chinese and overseas commercial systems, it is most feasible to integrate these four capabilities of XPeng: hardware capabilities, environmental perception, business logic, and emotional value. We consider the fusion of these four capabilities to be the most viable path for serving a wide range of industries commercially.
Of course, we will start with a select few industries. Simultaneously, we will open up our AI large model and, subsequently, our HDK (Hardware Development Kit) to support app development and ecosystem expansion. These are initiatives we plan to undertake in the next phase. Through commercial partnerships, XPeng's robots will leverage XPeng's overall distribution channels and create new online sales channels. Our goal is to directly connect with mass consumers and SMB clients through both online and offline integration. I believe this represents a highly valuable and unique commercial sales logic, distinct from other robotics companies. Thank you.
Thank you very much for your question. For X Points robots, yes indeed when it comes to commercialization, so we have gone down a different route versus other competitors, Many other competitors they are focusing on breaking into the market by ways of entering into factories, home usage and mainly for to be business.
Our strategy focuses on large-scale commercial scenarios (CS) as well as small and medium-sized business (SMB) applications. In other words, we are entering the market starting with commercialization cases, followed by industrial and household applications at a later stage, which will involve a broader range of SKUs.
We selected sales personnel and tour guides as initial target users because we believe our robots possess four core comprehensive capabilities that are highly beneficial and well-suited for these sectors, both in China and abroad. These four capabilities are: 1) the robot's main body and hardware; 2) environmental adaptability; 3) the services it provides; and 4) the emotional value it delivers.
Therefore, we are starting with commercial use cases, targeting smaller businesses and SMB industries. As the market opens up, we will also release SDKs to enable secondary development, further expanding user cases through collaborative commercial scenarios.
We believe this approach will open up more channels for Excellence Robots' business, covering both offline and online sales. This allows customers, ranging from large enterprises to SMBs, to see practical use cases for our robots. This strategy differentiates us from our peers.
Thank you, Mr. He Xiaopeng. My second question is about the company's self-developed dexterous hand. What is the latest progress? What overall design architecture has been adopted, and how does it compare to industry standards in terms of performance and cost?
The second question is about the latest progress on the company's self-developed dexterous hand. What overall design approach has been adopted, and how does it compare with peers in terms of performance and cost?
He Xiaopeng
Yes, I believe the dexterous hand is a critical component for robotics companies. It involves a set of hardware, sensing capabilities, and software algorithms. We have chosen to self-develop both the hardware and software. Fortunately, the development of our dexterous hand has proceeded very smoothly. We plan to begin mass production by the end of this year, alongside our complete robot units. Beyond R&D, we have also invested heavily in manufacturing processes and equipment for the dexterous hand.
Our dexterous hand uses direct drive technology, with 21 degrees of freedom per hand. Its size and thickness are comparable to an adult human hand. I believe our dexterous hand leads the industry in load-bearing and grasping capabilities. We did not adopt the tendon-driven hands widely discussed in the industry, as we do not consider them the optimal choice at present.
Therefore, in developing the hand, we did not pursue extreme force or precision alone. Instead, we focused on balanced capabilities, such as safety. We are even developing synthetic skin for the dexterous hand. We prioritized cost-effectiveness, reliability, maintainability (including quick-release mechanisms), and manufacturability for mass production. Our goal is to achieve scalable mass production based on this balanced approach. Thank you.
Thank you very much for your question. Indeed, the dexterous hand is an extremely important component of our robotic hands. For our robotic hands, we have developed one set of hardware, one set of software, and three different perception systems. Regarding specific mass production plans, which we will communicate to analysts by year-end, please note that we are not only conducting R&D on the dexterous hand but have also made significant investments in the manufacturing processes and equipment for its production.
In this regard, our robotic dexterous hand features 21 degrees of freedom, as mentioned earlier. In terms of size, it is identical to an adult human hand. With respect to load-bearing capacity, as well as grasping and gripping capabilities, we believe our dexterous hand holds a leading position compared to industry peers.
I am aware that the industry often discusses another type of robotic hand, but we do not consider that approach to be the optimal choice. Our focus is not solely on force or accuracy. Instead, we prioritize striking a strong balance among safety, reliability, ease of maintenance, and cost. For instance, our dexterous hand features a sophisticated skin layer that closely resembles human skin.
Operator
Your next question comes from Nikolay at JPMorgan. Please go ahead.
Nikolay
Okay, thank you. Let me ask a question regarding the ADAS LLA business. Mr. He Xiaopeng previously mentioned that the LLA business might list overseas in 2025. I would like to know which car models are initially planned for this, and whether the LLA 2.0 technology will be rolled out overseas before next year. Additionally, regarding the business model, it was mentioned that overseas markets would consider subscription and à la carte pricing. Could you share more details on this? Furthermore, are there opportunities in the medium to long term for our autonomous driving technology to collaborate with other OEMs overseas, thereby creating additional revenue streams and unlocking potential upside?
My first question relates to what Chairman He Xiaopeng mentioned: that in the first half of 2027, deployment of a model in overseas markets, particularly starting from Germany using VOA 2.1 technology, will begin. I wonder how many models will be equipped with this technology in overseas markets. Also, could you elaborate further on our business model, including subscription and one-time payment strategies for the long term in overseas markets? Thank you.
Hershel Peng
Thank you. Since the signal was not very clear, I could only catch the general gist, so I will provide a brief response. First, the L03 equipped with Turing chips has already been launched in overseas markets. Moving forward, all our subsequent vehicle releases will include an 'Ultra' version for overseas markets. This means that our hardware capabilities supporting VOA will cover the L03 and all future models and markets.
Our current progress is proceeding very smoothly, including regulatory preparations and the initiation of localized testing as just mentioned. We are indeed exploring a commercial software subscription business model, and we will announce details at the appropriate time. We are also establishing a relevant Business Development (BD) team. Currently, multiple partners are engaging with us to discuss how to globalize our VOA and other related capabilities.
Thank you very much for your question. Given that the signal was not coming through clearly and I could only hear some keywords, I will do my best to answer based on what I understood. First, regarding the Turing AI chips, they have already been deployed in our L03 vehicles, which have been launched in overseas markets.
All of these will feature the ultra version for VLA, which will be deployed in L03 models across all markets. Meanwhile, we are simultaneously advancing compliance with local laws and regulations, as well as conducting localization testing.
In the meantime, we are evaluating software subscription services for our customers. Details will be announced in due course. We have also established a new Business Development team that is actively discussing BLA usage with our partners, with potential expansion into other areas. Thank you.
Operator
Your next question comes from Tina Hu with Goldman Sachs. Please go ahead.
Tina Hu
Ah, thank you for the opportunity to ask a question. My first question is for Mr. Xiao Peng. As we ramp up mass production of humanoid robots—and congratulations on the initial financing round announced today—what is the outlook for profitability in the humanoid robot sector? Specifically, what is the expected timeline or required production scale to reach profitability? Additionally, does the company plan to separately report the profits of the humanoid robot business and the automotive business? This would provide investors and the market with clearer financial visibility into each segment.
So my first question is regarding the volume production of our humanoid robot product. Congratulations on the equity fundraising announced today. I am wondering what the expected timeline for profitability is for the humanoid robot business, or in other words, what sales volume needs to be achieved to become profitable? Related to that, do we have any plans to report profit levels separately for the humanoid robot and automotive businesses so that investors can better understand the profitability of these two distinct segments? Thank you.
Brian Group
Hey, Tina, this is Brian. Let me address your question regarding the financial outlook for the robotics business. I think it is a bit too early for us to comment. Currently, we are focused on the milestones previously shared, specifically achieving Start of Production (SOP) for volume manufacturing capability by the end of this year.
We also plan to start deploying the robots in internal scenarios first, then gradually offer them to external customers starting in the first half of next year, with a gradual ramp-up thereafter. That is our goal. Therefore, I would say it is too early for us to provide volume guidance.
We anticipate that the humanoid robot product will achieve significantly higher gross profit potential compared to the automotive business. In fact, the hardware margins are already much higher than those in the automotive sector. In addition, we see significant opportunities to generate further revenue from AI model training upgrades and software capabilities, which carry even higher margins.
Given the high probability, I would say the expectations are favorable, along with significantly lower investment and CapEx requirements for the robotics business. Once mass production volume is achieved, I believe this probability can materialize much faster than in the automotive business. That is our projection.
Regarding your question on the potential separation of the businesses, currently, they are operating together. In fact, we have not yet initiated any separation. As announced, there is an 18-month period that will gradually allow for the separation to be achieved.
In the meantime, our focus remains on achieving a high degree of synergy by leveraging capabilities in AI, advanced manufacturing, powertrain, supply chain, and so forth. This allows both businesses to achieve high efficiency and enhanced capabilities.
Therefore, in the near term, the expectation is that the businesses will still be viewed primarily as a combined entity. As mass production and commercialization scenarios become clearer, we will likely consider separation more seriously. However, given the ownership structure, this business will remain 100% consolidated.
This will not impact our financials going forward, even if the business begins to separate according to the plan. In short, we continue to consolidate all financials of the robotics business within the group, while simultaneously focusing on the most efficient and dynamic ways to operate the business. Thank you.
Operator
Thank you. That concludes the Q&A session. I would now like to turn the call back over to the company for closing remarks.
Alex C
Thank you once again for joining us today. If you have further questions, please feel free to contact our Investor Relations team via the contact information provided on our website for investor and financial communications. This concludes today's conference call. You may now disconnect your line. Thank you.
More details:XPeng IR
Disclaimer: The above content is generated by an AI language model based on public data and third-party automatic subtitles. The above content does not represent any position of Futu and does not constitute any investment advice. Futu Group makes no express or implied warranties or representations regarding the accuracy, timeliness, or completeness of the above content.
Risk Disclaimer: The above content only represents the author's view. It does not represent any position or investment advice of Futu. Futu makes no representation or warranty.Read more
Comments (149)
to post a comment
45
5