What else are the new car-making forces doing besides building cars?

Leapmotor 103,000 units, XPeng 39,000 units, NIO 36,000 units, Li Auto 38,000 units. These are the August sales figures of the four Chinese new EV manufacturers. The sales volume of the first three players remained basically flat compared with July, while Li Auto sold about 7,200 more units in August due to the low base in July.

When the ranking of new EV manufacturers gradually stabilizes, what excites the capital market is no longer monthly sales data, but what else these automakers can do beyond selling vehicles.

Recently, the four automakers have successively released their semi-annual reports, reflecting their respective ambitions and anxieties. XPeng and Leapmotor released their reports on the same day. During the earnings call, the topic analysts asked XPeng most frequently about was its robot business; while the management of Leapmotor spent a large portion of the meeting introducing the pace of its overseas expansion and collaboration with Stellantis.

Among the four, Li Auto has the largest cash reserve. It tries to emphasize its AI attributes beyond vehicle manufacturing, but analysts' questions eventually circled back to sales. NIO, which released its semi-annual report on September 1, continues to deepen its core vehicle manufacturing business relying on its multi-brand matrix.

The differentiation among new EV manufacturers has evolved from sales competition to the competition of developing the "second growth curve".

Different Earnings Call Scenarios

This is an atypical earnings call for an automotive company. If you did not know that this is XPeng Motors' session, you might think it is a robot company interpreting its financial report.

Right at the opening, analysts from investment banks including Morgan Stanley and Bank of America Merrill Lynch raised consecutive questions all related to XPeng's robot Iron: What is the delivery target for 2027? What is the unit cost of the mass-produced version? What is its gross margin? What kind of customers will buy Iron?

It is not unreasonable for analysts to deviate from the automotive business topic. On the day the financial report was released, XPeng's robot business officially announced the completion of a $900 million financing, with a post-investment valuation exceeding $6.3 billion. The investors include IDG, Gaorong Ventures, Tencent and Alibaba. Back in June, He Xiaopeng personally took the lead and served as CEO of the robot business, coordinating the collaboration of the complete vehicle, supply chain and AI teams.

During the call, He Xiaopeng stated that Iron is scheduled to start mass production at the end of 2026, and will first be deployed in XPeng stores for navigation and guidance services. It is expected to start external delivery in the first half of 2027, and the production capacity will be accelerated next year, which can be expanded to thousands of units per month or higher according to market demand.

Due to the high overlap of supply chains and the natural adaptation between autonomous driving and AI, automakers are considered players that are more likely to cross over to the embodied intelligence sector. At present, 85% of Iron's supply chain overlaps with that of XPeng Motors, and the most direct benefit is cost reduction. According to industry conventions, the pricing of Chinese robot products is usually 2.5-3 times the BOM cost. If AI models and software subscriptions are added, other sustainable revenue streams may be opened up. But at this stage, the application scenarios of humanoid robots are limited. "Compared with automobiles, the production capacity of robots will be much smaller," He Xiaopeng said.

Although the robot business took the center stage of the earnings call, the foundation of the business is still automobile manufacturing. In the first half of this year, XPeng's revenue reached 32.78 billion yuan, down 3.8% year on year; among which, affected by the decline in sales, the automotive sales revenue was 28.05 billion yuan, down 10.3% year on year, accounting for 85.6% of the total revenue.

In contrast, revenue from services and other segments was 4.73 billion yuan, up 67.1% year on year. This part of revenue comes from XPeng's technology R&D services provided to third parties such as Volkswagen on the one hand, and the sales revenue of auto spare parts on the other hand.

Although XPeng's automotive gross margin is 12.1%, the gross margin of services and other segments is as high as 71.4%, making the overall gross margin in the first half of the year 20.6%. The data is not bad, but XPeng's net loss in the first half of the year reached 3.12 billion yuan, 2.7 times that of the same period last year. The expansion of losses mainly comes from the expense side and continuous investment in new businesses such as robots and flying cars.

Using the cash flow from car sales to support the robot business sounds like a high-stakes bet, but XPeng's logic is self-consistent. The VLA visual motion model, perception algorithm, path planning, motor control and supply chain management capabilities accumulated from autonomous driving can be migrated to humanoid robots. Similarly, Tesla, Xiaomi, BYD and other companies are all deploying robot businesses, which has become an open fact in the industrial chain.

The difference lies in the pace and determination. XPeng is the only player among the new EV manufacturers that puts the robot business at the strategic level and is willing to spend a lot of time discussing it on the earnings call, actively shifting its valuation logic from automobile manufacturing to physical AI.

But the other side of the coin is risk. Robots are still far from large-scale commercialization and burn money fast. If the core automotive business fluctuates due to price wars or product rhythm issues, can the continuously cash-burning robot business still build up the "second growth curve"?

The Second Curve: Overseas Expansion

If XPeng's second growth curve is in the "future tense", Leapmotor's second growth curve is the "present tense" of overseas expansion. Leapmotor did not tell a sexy story like the robot business, and the high-frequency keywords on the earnings call were overseas expansion and collaboration.

Regarding overseas expansion, in the first half of this year, Leapmotor's export volume reached 96,300 units, a year-on-year increase of 372.6%, accounting for 27% of total sales, exceeding the total export volume of the whole year of 2025. Li Tengfei, Senior Vice President of Leapmotor, revealed that the company's original overseas sales target for this year was 100,000 to 150,000 units, which will be raised to about 200,000 units; the overseas sales target for 2027 is set at 350,000 to 400,000 units, nearly doubling compared with 2026.

As the competition in the domestic automotive market intensifies, overseas expansion has become Leapmotor's second growth curve to break the growth bottleneck. In the first half of the year, Leapmotor delivered 356,500 units, a year-on-year increase of 60.8%, ranking first among new EV manufacturers; sales in July and August further increased, exceeding 100,000 units for two consecutive months. Although this path is not as sci-fi as robots and AI, it has more stable cash flow and higher certainty.

Similar to Leapmotor, automakers including BYD, Geely, Chery and Great Wall also regard overseas expansion as an important growth pole. For example, Chery's total sales volume in the first half of the year was 1.3575 million units, of which 943,800 units were exports, accounting for nearly 70%; Geely's exports in the first half of the year increased by 158% year on year to 474,200 units, meanwhile, its exports in July exceeded 106,700 units, with a year-on-year growth rate of 202.4%.

Compared with these automakers, Leapmotor has a unique model for overseas expansion, that is, collaboration with Stellantis. Leveraging Stellantis' channel and brand advantages overseas, Leapmotor enters the overseas market with lower investment.

Especially on the manufacturing side, in recent years, Chinese automakers have successively built factories overseas to accelerate market expansion. Leapmotor also uses Stellantis' factories in Malaysia, Spain, Brazil and other regions to accelerate the layout of localized production, so as to reduce capital expenditure and time cost. Localized production can reduce tariff costs, but the procurement price of some parts is higher than that in China, and the comprehensive profit is not as significant as the market expects.

Especially in the initial stage of cooperation with Stellantis, the gross margin level of the agreed overseas models is relatively low, reflecting that Leapmotor's short-term priority for overseas expansion is to seize market share rather than pursue profitability. Leapmotor International achieved profitability in 2025. Although there was a small loss in the first half of 2026, the management explained that it mainly came from exchange gains and losses, which is expected to be covered in the second half of the year, and the whole year will still be profitable.

Leapmotor's model of "prioritizing scale over profit" will continue. During the earnings call, Leapmotor's management stated frankly that the current stage is a critical node for Chinese automakers to seize overseas markets. They will certainly attach importance to profitability, but will put sales growth in the most important position.

Among new EV manufacturers, Leapmotor has achieved profitability for three consecutive semi-annual periods, proving that relying on extreme cost-effectiveness and overseas expansion dividends is a more pragmatic choice for automakers.

Return to the Essence of Automobile Manufacturing

Among the four players, Li Auto is perhaps the one that most wants to prove it is "more than an automaker". Previously, Li Xiang has emphasized on many occasions that Li Auto is not just an automaker, but also an artificial intelligence enterprise. From self-developed chips and self-developed batteries to end-to-end intelligent driving, every move is moving in the direction of a technology company.

However, at the earnings call, the occasion that tests real profitability the most, the order of analysts' questions reveals the real attitude of the market. Unlike XPeng's Iron robot that was heatedly discussed, analysts first paid attention to Li Auto's sales volume and cost, and only then turned to AI at the end.

The first half of the earnings call was almost entirely about automobiles. In the second quarter, Li Auto delivered 98,300 units, down 11.5% year on year. Ma Donghui, President of Li Auto, attributed the reason to the model update and iteration. As Li Auto's extended-range L series enters the end of its product cycle, all models from L9 to L6 this year have been switched to the new generation platform. During the transition period of clearing inventory and giving profits for old models and ramping up production of new models, Li Auto's sales volume and gross margin are under pressure.

The extended-range technology used to be Li Auto's comfort zone, but as the extended-range track shrinks as a whole, Li Auto is also accelerating its transformation to pure electric vehicles. At present, the order structure of Li Auto's extended-range and pure electric vehicles is close to equilibrium. Li Auto's sales volume in July was about 30,300 units, of which the pure electric model i6 contributed about 15,300 units, almost accounting for half of the total.

However, the low price of i6 dragged down the profitability performance. In the first half of the year, Li Auto's vehicle gross margin dropped from 19.6% in the same period of last year to 7.8%, and the financial report frankly admitted that this was affected by the "different product mix". In other words, when the proportion of high gross margin extended-range vehicles decreases and the sales volume of cost-effective i6 increases, profits will inevitably be sacrificed.

Coupled with the price increase of raw materials such as chips and lithium carbonate, Li Auto recorded a loss of 3.98 billion yuan in the first half of the year, while it made a profit of 1.744 billion yuan in the same period last year.

If Li Auto is "forced" to focus on automobile manufacturing, then NIO has actively built a multi-brand matrix covering NIO, Onvo and Firefly, and continues to deepen its first growth curve of automobile manufacturing. In the second quarter, the sales volume of NIO, Onvo and Firefly was 61,000 units, 29,000 units and 18,000 units respectively, and the total sales volume of the three increased by 49.4% year on year. While the main brand maintains stable output, Onvo and Firefly enter different price bands, and the brand matrix has initially achieved sales growth.

However, the multi-brand layout has also increased marketing and R&D investment. NIO's net loss (GAAP basis) in the second quarter was 528 million yuan. Although the loss narrowed by nearly 90% year on year, it expanded compared with 332 million yuan in the first quarter (calculated on the basis of net loss attributable to common shareholders, the net loss in the second quarter was 722 million yuan). Although NIO's adjusted operating profit in the second quarter has turned positive to 210 million yuan and has achieved adjusted profitability for three consecutive quarters, it still takes time to achieve overall profitability on the GAAP basis.

In the second half of Li Auto and NIO's earnings calls, AI was mentioned. Different from XPeng that talked a lot about the IRON robot at the opening, Li Auto's introduction of AI was more closely aligned with its core automobile manufacturing business, focusing on the application of self-developed batteries, Mach M100 chips and VLA models on new models. NIO was only asked one question about external investment in AI, and AI played a marginal role in the earnings call.

The two founders have a more pragmatic attitude towards AI. In the view of Li Xiang, founder of Li Auto, batteries and chips are the most core barriers, and embodied intelligence is just a natural extension direction. The founder of NIO said more frankly that the company supports Ren Shaoqing, the head of the intelligent driving department, to set up an AI company outside, and NIO participates in it as a strategic shareholder. This can help NIO attract top talents from external strategic shareholders, but at present NIO still focuses on its core business.

For these two companies, what the market cares most about is still fundamentals such as whether sales can be stabilized, gross margin level, and market performance of new products. AI is more like a slow variable that has quietly occurred, but it has not yet reached the stage where it can independently support the valuation story.

Judging from the current results, no matter XPeng bets on humanoid robots, Leapmotor goes all in on overseas expansion, Li Auto moves closer to AI, or NIO deepens multi-brand automobile manufacturing, none of the second growth curves can be said to have been fully validated globally. But it is certain that the growth logic that only relies on selling cars in the single domestic market has come to an end. New EV manufacturers must find new incremental space. Whoever can convert the second growth curve into stable cash flow first is more likely to take the initiative in the next round of competition.

This article is from the WeChat official account Leopard Change (ID: baobiannews), written by Chen Fashan, published with authorization from 36Kr.