Xiaomi’s $10 Billion EV Bet: Can Lei Jun Scale a Car Business and Keep Buyers’ Trust?

In the second quarter of 2026, Xiaomi delivered 104,199 vehicles, 28.2% more than a year earlier. During the same period, the business segment that includes its cars, artificial intelligence work and other new initiatives recorded an operating loss of RMB 2.6 billion.

The contrast captures the challenge facing Xiaomi’s electric-vehicle ambitions. Customers are buying its cars, but growing deliveries have not guaranteed consistent operating profits. At the same time, the company is dealing with safety recalls and the responsibilities that come with building vehicles rather than consumer electronics.

When Xiaomi committed up to US$10 billion to its electric-vehicle business in 2021, founder Lei Jun presented the project as a defining entrepreneurial undertaking. The early commercial response suggested that a smartphone maker could compete in China’s crowded car market. The harder question now is whether Xiaomi can turn that response into a business that earns sustainable returns while meeting the safety expectations placed on every automaker.

The evidence points to a qualified answer. Xiaomi has demonstrated that it can attract buyers and expand production quickly. It has not yet demonstrated that strong demand will consistently translate into operating profit, nor has it finished proving how its approach to product safety will hold up as its fleet grows.

The decision behind the $10 billion commitment

On March 30, 2021, Xiaomi announced through a Hong Kong Stock Exchange filing that its board had resolved to establish a smart electric-vehicle business. The plan called for an initial investment of RMB 10 billion, with total investment expected to reach approximately US$10 billion over the following decade. Lei Jun would lead the new business as chief executive.

The commitment was substantial, but Xiaomi was not starting from the position of a cash-strapped startup. The company reported RMB 108 billion in cash at the end of 2020. Lei later said Xiaomi intended to finance the venture with its own resources.

In a speech in July 2024, Lei offered a personal account of the deliberations behind the decision. He recalled that the process began in January 2021 and described extensive research, discussions with industry participants and arguments for and against entering the automotive market. According to his account, the research involved more than 70 days of work, visits to multiple cities and discussions with hundreds of people.

That account provides useful context, but it should be treated as the founder’s retrospective explanation rather than proof that any single event determined the board’s decision. The formal filing establishes the commitment and its scale; it does not establish one exclusive motive for making it.

Xiaomi’s strategic logic was understandable. It already had experience designing hardware, integrating software and connecting devices through a common ecosystem. A car could extend that ecosystem into a product people use every day. Yet the move also exposed Xiaomi to a very different set of responsibilities.

A smartphone can frustrate a customer when software fails or a component breaks. A car has to operate reliably at speed, protect occupants in a collision and provide clear warnings when its driver-assistance systems reach their limits. The business therefore required more than adapting consumer-electronics expertise to a new product. It required Xiaomi to build the engineering, manufacturing, service and safety capabilities of an automaker.

The SU7 sedan, which went on sale in March 2024, gave the company its first major test of that strategy. Its competitive pricing and Xiaomi’s established brand helped attract attention in a market where local manufacturers were already fighting hard for customers. The response showed that buyers were willing to consider a newcomer with a background in phones and connected devices.

That was a meaningful first result. But a successful launch tests demand, not the entire business model.

What the financial results reveal

Xiaomi’s 2025 results provided the strongest evidence yet that the EV strategy could become commercially significant. The company delivered 411,082 vehicles during the year, an increase of 200.4% from 2024. Revenue from its smart EV, AI and other new initiatives segment reached approximately RMB 106.1 billion, up 223.8%. Vehicle revenue accounted for RMB 103.3 billion of that total.

The segment’s gross margin rose to 24.3%, compared with 18.5% in 2024. It also reported an operating profit of RMB 0.9 billion for the full year, its first annual operating profit.

Those numbers marked a clear advance from the investment phase. Xiaomi had moved beyond selling a small number of vehicles to generating substantial revenue from a growing operation. It had also shown that the segment could report an operating profit over a full year.

However, the size of that profit matters as much as the milestone. RMB 0.9 billion against RMB 106.1 billion in revenue represents a narrow operating margin. It leaves relatively little room for cost increases, pricing pressure or an expensive period of expansion.

There is another accounting distinction worth keeping clear: Xiaomi reports its cars together with AI and other new initiatives. The segment’s operating result is not a standalone profit figure for the automotive business. It should not be presented as though it measures the financial performance of the cars alone.

The quarterly figures also showed that progress was not a straight line. Segment gross margin declined from 25.5% in the third quarter of 2025 to 22.7% in the fourth. The full-year result was positive, but the final quarter already suggested that sustaining margins could be difficult.

The pressure became more visible in 2026. The segment recorded an operating loss of RMB 3.1 billion in the first quarter and RMB 2.6 billion in the second. Its second-quarter gross margin fell to 19.2%, from 26.4% a year earlier.

Xiaomi attributed the margin decline to a lower contribution from deliveries of its higher-priced SU7 Ultra, increased prices for key components and higher costs associated with its AI business. Segment operating expenses reached RMB 7.4 billion in the second quarter, an increase of 25.7% year over year.

These details matter because they complicate the easy assumption that selling more vehicles will automatically make the business more profitable. A growing manufacturer must manage product mix, component costs, development spending and the expenses required to support a larger customer base. If those costs rise faster than the value generated by each additional vehicle, volume alone will not solve the problem.

The wider Chinese market makes Xiaomi’s delivery performance notable. Xiaomi reported 28.2% year-over-year growth in second-quarter deliveries while China’s overall passenger-vehicle retail market declined by about 22%, according to the industry figures cited in its interim results. Xiaomi was therefore expanding against a difficult market backdrop.

Yet the company’s wider financial position should not be confused with the performance of its EV-related segment. Xiaomi Group reported second-quarter revenue of RMB 108.9 billion and adjusted net profit of RMB 6.2 billion. Adjusted net profit was down 42.6% year over year, with smartphone profitability also under pressure from higher memory-chip costs.

The group remained profitable, but the figures show why management cannot judge the automotive project by deliveries alone. The segment must become more resilient financially while competing for capital and management attention with Xiaomi’s established businesses and its other technology investments.

Safety became a test of the business

The automotive strategy also faced a challenge that revenue figures cannot capture.

On March 29, 2025, a Xiaomi SU7 Standard crashed on a highway in Anhui province and caught fire, killing three people. In a statement reported by Reuters, Xiaomi said the vehicle had been using its navigation-assistance system before the driver took control. The company said it had provided vehicle and system data to the police and would cooperate with the investigation.

The circumstances of a fatal crash must be handled carefully. A reported sequence of events is not, by itself, a complete finding about responsibility. Nor does a single accident establish that every vehicle of the same model is unsafe. But an incident involving a driver-assistance system raises serious questions about system limitations, driver warnings and whether customers understand what the technology can and cannot do.

Xiaomi’s recall history makes the broader product-governance issue more concrete. In January 2025, the company recalled 30,931 SU7 Standard vehicles because a software strategy issue could affect the intelligent parking-assistance system’s ability to detect stationary obstacles. Xiaomi said it would provide a free over-the-air software update.

In September 2025, the company announced a separate recall covering 116,887 SU7 Standard vehicles. China’s market regulator said the Level 2 highway-assistance function might not identify, warn about or handle certain extreme scenarios adequately. If a driver failed to intervene promptly, the situation could increase collision risk. Xiaomi again planned a free software update.

The distinction is important: these notices identified specific risks affecting defined vehicle populations. They do not establish that every recalled vehicle experienced a failure, or that the recalls caused any particular crash. They do, however, show that software reliability and the limits of driver assistance are central responsibilities for Xiaomi’s automotive business.

Door releases became another safety concern. In August 2026, China’s market regulator announced recall campaigns involving 4,275,743 vehicles from nine manufacturers over concerns that interior mechanical emergency releases could be difficult to identify or operate. Xiaomi’s portion covered 390,435 SU7 vehicles. The announced remedies included warning labels and software changes intended to improve unlocking logic and window-lowering behaviour after a crash.

The wider industry context matters. The regulator’s action was not directed at Xiaomi alone, and Tesla accounted for the largest share of the vehicles included in the broader recall announcement. Nevertheless, a vehicle’s emergency-release mechanism is not a minor design detail. In a serious collision, occupants or rescuers may need to open a door when normal electrical systems are no longer working.

Taken together, these campaigns do not prove that Xiaomi’s cars are broadly unsafe. They do demonstrate why a fast-growing automaker must treat safety as a continuing engineering and governance obligation, rather than a matter to address only after public criticism.

There is also no basis in the cited public financial results to claim that these recalls caused Xiaomi’s operating losses in 2026. The financial and safety issues should be assessed separately. They are connected by management responsibility, but a direct financial cause-and-effect relationship has not been established.

Trust depends on what changes inside the company

For Xiaomi, the next question is not simply how it responds to criticism. It is whether lessons from product problems become lasting changes in the way vehicles are designed, tested and supported.

A software update can address a particular defect. A recall can warn owners and reduce a defined risk. But the deeper test is whether a company’s processes are designed to identify problems before vehicles reach customers, communicate limitations clearly and give safety concerns sufficient weight when commercial deadlines are pressing.

In February 2026, Lei Jun publicly discussed a stronger focus on vehicle safety, including plans to involve outside experts. Reports also described changes to Xiaomi’s internal safety-review arrangements. Such measures may help, but their existence alone cannot establish how effective they are. The meaningful evidence will be how the processes operate, what problems they uncover and whether they lead to timely corrective action.

That distinction also matters for communication. When a company faces a safety controversy, defending its reputation may be understandable, but it cannot substitute for clear explanations of what happened, what remains unknown and what is being changed. Customers need information that helps them understand the limits of a system and the actions required of them.

Whether Xiaomi has fully restored buyer trust cannot be determined from sales numbers alone. Continued demand shows that customers are still purchasing its vehicles. It does not reveal how every owner feels, whether a buyer would recommend the brand after a difficult experience or how the incidents will affect future purchasing decisions.

Trust is therefore better treated as an outcome to measure over time than as a claim a company can make about itself. Repeat purchases, customer retention, service experience, transparent safety reporting and the effectiveness of corrective actions will all provide more meaningful evidence than a single month of orders.

Can the delivery target support a sustainable business?

Xiaomi’s ambition also depends on whether it can maintain its pace of expansion. The company set a 2026 delivery target of 550,000 vehicles. By the end of September, reported monthly deliveries had reached 40,015 for that month, following 30,153 in August.

Using the reported January-to-July total of 216,322 vehicles, the August figure and September deliveries produces a cumulative total of approximately 286,500. On that basis, Xiaomi would need to deliver about 263,500 vehicles in the fourth quarter to reach 550,000—roughly 88,000 a month.

That is a demanding pace. For comparison, Xiaomi delivered 145,115 vehicles during the whole of the fourth quarter of 2025. The comparison does not prove the 2026 target will be missed; production capacity, new models, customer demand and delivery timing can change. It does show that the remaining requirement is materially higher than Xiaomi’s recent monthly run rate.

The calculation should also be understood for what it is: an estimate based on reported deliveries, not an announced company forecast for each remaining month. October delivery data and the next financial results will provide a clearer picture.

Even if Xiaomi reaches its target, the number will not settle the larger question. Rapid expansion can improve manufacturing efficiency and spread fixed costs over more vehicles. It can also increase pressure on factories, service networks, quality control and the systems used to investigate defects. The value of growth depends on whether the company can manage those demands without allowing costs or safety risks to grow faster than the business.

What the bet asks of Lei Jun

Xiaomi’s automotive expansion is a bet that strengths developed in consumer electronics can support a much more complex product. The company has demonstrated an ability to attract customers, launch vehicles at scale and generate substantial segment revenue. Its 2025 operating profit showed that the business could cross an important financial threshold.

But the 2026 losses show that the threshold has not yet become a stable floor. The segment’s profitability remains sensitive to product mix, component prices, AI spending and operating costs. The recall history shows a parallel challenge: building cars requires safety systems, clear communication and governance that work reliably as production expands.

Xiaomi is also investing in research and development across AI, chips, intelligent driving and other areas. Its 2025 research and development spending was RMB 33.1 billion, and the company has indicated that it expects to invest more than RMB 200 billion across 2026–2030. That is a company-wide commitment, not an EV-only budget. It illustrates the competition for resources that management must balance as it develops several technology businesses at once.

Three lessons emerge from Xiaomi’s experience.

First, growth and profitability must be evaluated separately. Deliveries can rise while margins fall, so leaders need to understand what each additional sale contributes after costs.

Second, safety is a product and governance responsibility. Recalls and software updates can address defined issues, but lasting confidence depends on how consistently a company identifies risks, explains limitations and acts on evidence.

Third, a founder’s personal commitment can help a company move quickly, but it also raises the stakes. Lei Jun’s reputation is closely tied to the project. The strongest evidence of leadership will not be the scale of the original promise; it will be the quality of the operating system, safety culture and financial discipline built to fulfil it.

Xiaomi’s EV decision has passed its first major test: customers have shown that they are willing to buy the cars. Its next test is more demanding. The company must demonstrate that it can keep growing, protect customers and earn a durable return without treating any one of those goals as a substitute for the others. The next delivery report will show how quickly the business is expanding. The more important evidence will come from the margins, the response to safety problems and the consistency of results over time.