Introduction: XPeng is expanding its business boundaries, but new ventures have yet to generate stable returns.
November 13, London.
XPeng held a new vehicle launch event in the UK, introducing the G9L and MONA L03 models to the British market. Simultaneously, the company announced plans to launch five right-hand drive models in the UK over the next year and establish 100 retail and service outlets by 2027.
This was not merely an overseas car-selling initiative.
XPeng aims to build its own sales and service system in the UK and plans to bring intelligent driving, humanoid robots, and Robotaxi businesses to the local market. According to the company's roadmap, XPeng aspires to evolve beyond being just an automobile manufacturer into a global enterprise capable of exporting AI technology.
A few days later, the company released its third-quarter financial results.
The figures did not fully align with market expectations for a rapidly expanding enterprise. XPeng reported Q3 revenue of CNY 24.28 billion, a 19.1% year-on-year increase, but net loss attributable to shareholders reached CNY 868 million, widening by approximately 128% compared to the same period last year.
Revenue hit a single-quarter high, while losses expanded.
Behind these numbers lies a transformation XPeng is undergoing: competing for sales volume in China's NEV (New Energy Vehicle) market, selling technology to other automakers, and investing heavily in autonomous driving, robotics, and overseas markets.
The company wants to do more things, but these new businesses are still some distance away from forming stable revenue streams.
First, Solving Its Own Market Position
In recent years, China's NEV market has undergone significant changes.
The sector has shifted from competition among a few brands for technology and market share to a stage where numerous automakers compete simultaneously. BYD, Geely, Huawei-affiliated vehicles, Xiaomi Auto, NIO, Li Auto, and others are all vying for consumers across different price segments.
Early on, XPeng primarily built brand recognition through intelligent driving and smart cockpits. However, in the NEV market, technologies that once created differentiation are gradually becoming standard features across an increasing number of models.
When purchasing cars, consumers now consider not only driver-assistance functions but also price, space, range, charging speed, brand, and after-sales service.
This has forced XPeng to adjust its product strategy.
The company's current product layout has expanded significantly. The MONA series targets the mass market and bears the burden of sales volume, while the G series and G9L attempt to enter higher price brackets, and models like the X9 cover further niche segments.
This shift is already reflected in delivery data.
In September this year, the MONA L03 delivered over 10,000 units in a single month; the XPeng GX delivered 7,224 units. The G9L, launched domestically on September 17, has a limited-time starting price of CNY 231,800.
On October 29, XPeng introduced the MONA L05 with a limited-time starting price of CNY 139,800, with deliveries beginning on November 1.
From a product perspective, XPeng is attempting to simultaneously cover markets ranging from CNY100,000+ to CNY 300,000+.
However, an expanded product line also means the company must bear higher costs.
Different models require independent product development, supply chain management, production scheduling, and marketing resources. Especially amid intense domestic price competition, launching new cars does not necessarily mean the company can achieve higher profits.
XPeng delivered 118,400 vehicles in Q3, a 2.1% year-on-year increase, while revenue grew by 19.1%. Improved product mix and higher average selling prices indeed drove revenue growth.
Yet, the gross margin for the automotive business remained at 13.1%, flat compared to the same period last year.
This indicates that while XPeng has managed to boost revenue through product adjustments, it has not yet sufficiently proven that it can generate higher profits relying on these products.
In October, XPeng's deliveries reached 50,237 units, a 20% year-on-year increase, setting a monthly record. Among these, overseas deliveries totaled 13,326 units, surging 166% year-on-year.
Sales volume is recovering, but what truly needs observation is which models contribute to this new volume and how much profit they leave for the company.
For XPeng, the past challenge was proving it could sell cars; the current challenge is proving that selling cars can be a profitable business.
Starting to Sell Technology to Others
If one views XPeng solely as an automobile manufacturer, it is difficult to understand why it continues to invest heavily in chip development, intelligent driving systems, and underlying software.
A key reason is Volkswagen.
In 2023, Volkswagen invested approximately USD 700 million in XPeng, acquiring about a 5% stake. The two companies subsequently announced a strategic partnership to jointly develop intelligent electric vehicles for the Chinese market.
The significance of this deal goes beyond Volkswagen becoming a shareholder of XPeng.
At the time, Volkswagen faced pressure regarding product development speed and intelligence capabilities in China's NEV market, while XPeng possessed a relatively complete intelligent vehicle technology system. Both parties hoped to combine Volkswagen's manufacturing, engineering, and quality control capabilities with XPeng's EV platform and software technologies.
In March 2026, the first model co-developed by both parties, the ID.UNYX 08, entered mass production at the Volkswagen Anhui factory.
Volkswagen stated that the timeline from the start of cooperation to mass production took only 24 months. The vehicle features an 800V high-voltage platform and supports Level 2 driver assistance and OTA upgrades.
This collaboration gave XPeng an opportunity to prove that its technology is not only usable in its own cars but can also enter the product systems of traditional global automakers.
The China Electronic Architecture (CEA) co-developed by Volkswagen and XPeng is also an extension of this cooperation. According to Volkswagen's plans, this architecture will be used in more models targeting the Chinese market in the future.
For XPeng, this may represent a business with greater imaginative potential than simply selling cars.
Automobile manufacturing requires bearing heavy costs related to production, inventory, channels, and after-sales services, whereas technology licensing and software services offer opportunities for different revenue models. If XPeng can provide intelligent driving, electronic architectures, and software technologies to multiple automotive brands, it will not need to rely entirely on its own vehicle sales to expand its business scale.
However, there remains a gap between technical cooperation and commercial success.
Reuters reported on October 20 that the ID.UNYX 08, co-developed by Volkswagen and XPeng, sold an average of about 267 units per month from April to August this year, far below the internal target of 3,622 units per month set by Volkswagen.
This demonstrates that shortening development cycles and introducing Chinese intelligent technologies does not guarantee consumer acceptance.
Technical cooperation can solve issues related to automotive development efficiency, but it cannot replace brand building, product positioning, and marketing.
XPeng still needs to prove that the technology it exports can create genuine commercial value, rather than merely serving as a way for partners to reduce R&D costs.
He Xiaopeng clearly hopes to scale this effort.
In October this year, at the Paris Motor Summit, he proposed "XPENG AI Engineering," aiming to open up a series of technological capabilities—from vehicle platforms and electronic architectures to AI chips and physical world models—to global partners.
This is essentially an attempt to extend XPeng's business boundaries from automobile manufacturing to technology supply.
If this system can be established, XPeng's future revenue sources may no longer be limited to whole-vehicle sales.
But on this path, it still needs to find enough paying customers willing to buy in.
Overseas Markets Becoming the Second Growth Curve
XPeng is accelerating its overseas expansion.
In October this year, the company's overseas deliveries reached 13,326 units, a 166% year-on-year increase, setting a monthly record.
As of November 13, XPeng stated that its sales network covers 83 countries and regions globally.
The latest moves in the UK market are a microcosm of this strategy.
XPeng launched the G9L and MONA L03 in the UK, with starting prices of GBP 64,990 for the G9L and GBP 32,990 for the MONA L03. Compared to the Chinese market, prices in the UK are significantly higher.
However, the price difference between domestic and international markets should not be simply interpreted as higher profits overseas. Local taxes, transportation, certification, channel costs, and product configurations all impact final profitability.
What truly deserves attention is that XPeng has begun building its own sales and service networks overseas.
According to the plan announced on November 13, the company will launch five right-hand drive models in the UK over the next year, cooperate with local retail and leasing enterprises, and aim to establish 100 retail and service outlets by 2027.
This means XPeng is no longer content with exporting cars via distributors but hopes to directly participate in local market operations.
This shift has practical reasons.
Competition in China's NEV market is intensifying, with more automakers viewing overseas markets as new growth sources. As export volumes increase, competition among Chinese brands is also emerging in overseas markets.
Relying solely on exports makes companies vulnerable to tariffs, logistics, and the capabilities of local dealers.
Building local sales, after-sales, and service systems helps automakers grasp customer needs more directly and fosters long-term brand recognition.
But the cost is increased investment.
Stores, after-sales services, spare parts supply, employee training, and local marketing all require capital. Especially in Europe, differences in regulations, driving habits, and consumer preferences across countries necessitate continuous localization adjustments.
XPeng has previously utilized contract manufacturing at the Magna Steyr plant in Austria for certain models to address EU tariffs and overseas delivery demands. As export scales grow, the company is considering further increasing local production capacity in Europe.
Globalization has brought new market space to XPeng, transforming it from an automaker primarily operating in China into a multinational entity managing markets across multiple countries.
Sales growth is only the first step. Whether local sales networks can generate sufficient order density, whether factory utilization rates can improve, and whether after-sales services can cover wider areas—these factors ultimately determine how much profit the overseas business retains.
But from a risk perspective, suppliers are also beginning to bear part of the risks that should originally be borne by customers, financing institutions, and infrastructure investors.
Seeking Another Identity for Itself
If overseas markets still revolve around automobiles, then humanoid robots and Robotaxis represent XPeng's more aggressive attempts.
On September 8 this year, XPeng announced the activation of the production line for its humanoid robot, IRON.
According to disclosed information, the automation rate of core processes on this production line exceeds 80%. IRON possesses 76 degrees of freedom in its body and 21 degrees of freedom in each hand, equipped with three XPeng Turing AI chips, achieving a theoretical effective computing power of up to 2,250 TOPS.
Behind these parameters lies XPeng's attempt to transfer capabilities accumulated in the automotive industry to robot manufacturing.
Automakers possess experience in large-scale production, precision manufacturing, supply chain management, and quality control, while humanoid robots similarly require motors, sensors, control systems, chips, and numerous precision components.
From a technical pathway perspective, XPeng hopes to allow automobiles and robots to share some underlying technologies, including AI chips, physical world models, and motion control systems.
XPeng has already introduced external capital into its robotics business.
In Q2 this year, the company disclosed that its robotics business completed financing exceeding USD 900 million, with a post-money valuation surpassing USD 6.3 billion.
This indicates that the capital market is willing to fund this business, but valuation does not equal commercial revenue, nor does it imply that robotics has formed a mature profit model.
XPeng currently plans to first deploy IRON in its own stores and campuses before gradually expanding to external clients. The company previously targeted scaled mass production by the end of 2026 and formal deliveries to domestic and international clients in 2027.
There remain many questions needing verification.
Can robots perform stable work in real-world environments? Can they replace a portion of human labor? What are the procurement and maintenance costs? How much are corporate clients willing to pay?
Before these questions are answered, humanoid robots remain a business requiring continuous investment.
Robotaxis face similar challenges.
On October 26, XPeng obtained permission to conduct Level 4 autonomous driving demonstration operations in Guangzhou. This means the company can carry out more advanced Robotaxi demonstrations locally.
Previously, XPeng had advanced its autonomous driving technology upgrade through VLA 2.0 and conducted relevant tests in Guangzhou.
However, demonstration operations are still distant from true commercialization.
Robotaxis require solving not only autonomous driving technology but also vehicle operation, insurance, liability assignment, dispatch efficiency, and unit economics.
Even if technology reaches a certain level, whether it can operate continuously without safety drivers and provide services at reasonable costs still requires time to verify.
For XPeng, the greatest value of robots and Robotaxis may temporarily lie not in direct profit contribution, but in providing long-term growth possibilities beyond the automotive business.
But this raises another question: When the automotive business has not yet achieved stable profitability, does the company have sufficient funds and management capacity to advance so many new businesses simultaneously?
What Truly Needs Proving Is Not How Many Things It Can Do
Returning to the Q3 financial report.
XPeng reported revenue of CNY 24.28 billion, a 19.1% year-on-year increase; net loss attributable to shareholders was CNY 868 million. The automotive business gross margin was 13.1%, and the overall gross margin was 18.9%.
The company's R&D expenses reached CNY 2.853 billion, and sales and administrative expenses amounted to CNY 2.776 billion. These two expense categories combined exceeded CNY 5.6 billion, surpassing the quarterly gross profit of CNY 4.6 billion.
However, the widening losses cannot be entirely attributed to the main business.
Operating losses in Q3 were approximately CNY 752 million, roughly flat compared to the same period last year. The expansion of net losses was also influenced by long-term investment losses and exchange rate losses.
This implies that XPeng's current issue is not a sudden loss of control in its automotive business, but rather that the main business has not yet formed sufficient profit margins, while the company is paying substantial costs for future expansion.
As of the end of September, XPeng disclosed cash reserves of CNY 45.41 billion, an increase from CNY 40.48 billion at the end of Q2. The company does not lack funds to support R&D and expansion in the short term.
But ample funding does not equate to unlimited indefinite investment.
The NEV industry has entered a new competitive phase. In the past, companies could gain market share through continuous financing, launching new models, and expanding sales. Now, investors are paying more attention to gross margins, operating cash flow, returns on R&D investment, and whether new businesses can form genuine revenue.
XPeng hopes to establish three growth curves: automobiles, robotics, and globalization.
The automotive business provides scale, globalization brings new markets, and robotics and autonomous driving offer longer-term technological commercialization opportunities.
This strategy has logic, but the three curves do not mature simultaneously.
The automotive business still needs to resolve profit margin issues; overseas business requires continuous investment in channels and service systems; and robotics and Robotaxis still need to prove commercial demand.
XPeng's dilemma is that all these businesses require funds but may not generate returns at the same time.
For Q4, the company expects deliveries of 169,000 to 174,000 vehicles, with revenue reaching CNY 30.4 billion to CNY 31.8 billion. Single-month deliveries in October already reached 50,237 units, hitting a historic high.
Sales growth provides the foundation for continued transformation, but the test for the next stage is no longer just sales volume.
Whether Volkswagen will continue purchasing XPeng's technology, whether overseas stores can form stable sales and service income, and whether IRON can find commercial clients truly willing to pay—these questions ultimately return to one metric: profit.
XPeng is trying to shed its identity as a mere automobile manufacturer.
But before becoming a global Physical AI company, it first needs to prove that it can master the business of making cars well.
After all, technology can bring new stories, and the market is willing to pay a premium for the future. But for a listed company, ultimately, someone must pay for these technologies and products.