Revenue exceeds 10 billion, but net profit shrinks 85%, Li Auto enters the "most dangerous year"

When it comes to stock trading, rely on Golden Kylin Analysts’ reports—authoritative, professional, timely, comprehensive—helping you uncover potential thematic opportunities!

(Source: Titanium Media APP)

Image generated by AI

In March 2026, Li Auto delivered a performance that many investors found shocking.

In 2025, Li Auto’s revenue was 112.3 billion yuan, a sharp 22.3% decrease year-over-year; net profit was 1.14 billion yuan, an 85.8% drop from 8.05 billion yuan in the previous year. Full-year deliveries shrank from 500,500 units in 2024 to 406,300 units, a decline of over 18%.

Even more concerning was the Q1 guidance: expected revenue of 20.4 to 21.6 billion yuan, well below market expectations of 24 billion yuan. This suggests Q1 year-over-year decline could continue, ranging from 16.7% to 21.3%.

2025 was the most challenging and transformative year for Li Auto since its founding. The company experienced the pain of losing sales dominance and a sharp profit plunge, while decisively initiating a deep organizational restructuring to shift from a “professional manager model” back to a “startup model.”

During the earnings call that evening, CEO Li Xiang spent much more time discussing organizational change, AI strategy, in-house chip development—and a recurring term: “embodied intelligence.”

A “triumphant defeat” with hundreds of billions in revenue: profits sustained by interest, deliveries falling out of top three

In 2025, Li Auto’s full-year revenue was 112.3 billion yuan, down 22.3%. This figure would be alarming for any company, but among new forces, it preserved the record of “three consecutive years of over 100 billion yuan in revenue.” Net profit was 1.14 billion yuan, a collapse of 85.8% from 11.8 billion yuan in 2024.

What’s more painful is that this 1.14 billion yuan was mainly supported by interest income and fair value changes in investments; operational losses amounted to 521 million yuan. According to financial data, the company’s pre-tax profit in 2025 was approximately 1.297 billion yuan, while net interest and investment income reached 1.919 billion yuan—these gains directly offset the operating losses from core business, enabling a positive net profit.

However, the capital market has not completely abandoned Li Auto, because another set of data is more significant: as of the end of 2025, cash reserves reached 101.2 billion yuan, ranking first among Chinese new forces. Free cash flow turned positive in Q4 at 2.47 billion yuan, indicating the toughest quarter is behind, and the company still holds a billion-yuan war chest.

The decline in deliveries directly explains the revenue shrinkage. In 2025, deliveries totaled 406,300 units, down 18.8% year-over-year, dropping Li Auto from the top spot among new forces in 2024 to fifth place—overtaken by HarmonyOS Smart Mobility, Leapmotor, Xpeng, and Xiaomi. The pure electric model MEGA faced difficulties, and competitors in the extended-range segment all arrived, marking Li Auto’s first experience of being “surrounded on all sides.”

Gross margin was one of the few bright spots in the financial report. The full-year gross margin was 18.7%, only narrowing by 1.8 percentage points from 20.5% in 2024, and in Q4 it even rebounded to 17.8%. The L6 model lowered the average selling price, but scale effects and cost control maintained the basic margin—meaning the price war has not yet broken Li Auto’s bottom line.

But the core issue remains unchanged: the L series faces a generational upgrade, and pure electric models are still climbing. The good news is that the pure electric i6 broke through capacity bottlenecks in early 2026, with a production capacity of 20,000 units in March, and orders for the i8 in March increased by 180% compared to January.

Li Xiang set a target for 2026 of over 20% year-over-year growth—based on 406,000 units, aiming for 490,000 units for the full year. The Q1 guidance of only 85,000–90,000 units means a strong rebound is needed in the second half.

The key to this rebound is the upcoming launch of the all-new Li Auto L9 in Q2. The financials have already set the bottom line: 100 billion yuan in cash is a moat, three years of profitability is confidence, but falling out of the top three and halving profits mean Li Auto cannot afford any mistakes.

Deny plans to close 100 stores, aiming to rebuild the sales system

More than the financial figures, what’s more noteworthy is the deep transformation Li Auto is undergoing. Before the financial report was released, rumors circulated that Li Auto planned to close 100 stores. Ma Donghui directly denied this during the call: “That is not true information.”

But denial doesn’t mean no action. In fact, Li Auto is reforming its sales system.

The core issue was openly addressed by Li Xiang himself: “Our biggest problem in the past was managing the direct-sales system with a dealer management approach.”

The core of direct sales is to operate through stores, but Li Auto’s past assessment mechanisms, incentive models, and site selection decisions were fundamentally still dealer-like—headquarters made decisions, store managers just executed.

On March 1, Li Auto officially launched the “Store Partner” mechanism. This is not franchising but delegating operational decision-making and profit-sharing rights to store managers, making them truly responsible for their store’s performance. New store site evaluations will involve store managers fully, with responsibilities and rights linked to personnel.

Ma Donghui explained straightforwardly: “We want to fundamentally solve the issues of blind store expansion and external store outreach.”

The goal of this mechanism is to cultivate a large number of store managers earning over one million yuan annually, with top performers earning two to three times the industry average. The effectiveness of this assessment is expected to be clearly visible by Q3.

Since 2025, several senior executives have left Li Auto, including Xia Zhongpu, responsible for intelligent driving end-to-end models; Jia Peng, head of technical R&D; and Wang Jiajia, head of mass production R&D. Some have even started their own ventures. Li Xiang expressed congratulations but also highlighted a key data point: many core talents cultivated over the past 5–10 years have gone on to start businesses and gained recognition in the capital markets; internally, a large number of post-90s and post-95s now hold key positions, and post-00s campus recruits are becoming core technical staff.

“This is the most important sign of our confidence in the next decade,” he said.

Half of R&D is focused on AI, with in-house chip “Mahe 100” about to go into mass production

If channel reform is defensive, then AI and chips are Li Auto’s offensive.

In 2025, R&D investment was 11.3 billion yuan, with 50% allocated to AI. In 2026, this investment will remain around 12 billion yuan, with AI still accounting for about half.

Where is the money going? The most significant is the in-house chip “Mahe 100.”

This 5nm process automotive-grade chip will be installed in the new Li Auto L9. During the call, Li Auto provided several key data points:

The BOM cost per Mahe 100 chip is significantly lower than third-party solutions, and by replacing the previous-generation XCU controller with Mahe 100, each vehicle can save over 1,000 yuan.

More importantly, Mahe 100 integrates deeply with the self-developed Xinghuan OS and full-line chassis control. The hardware-software integration gap, according to CTO Xie Yan, will be like the difference between Apple and Android—once established, it will create a structural and continuously expanding advantage.

Li Xiang explained this transformation from a technical philosophy perspective: upgrading from a “2D camera + lidar” solution to 3D ViT, enabling the vehicle to perceive and understand the physical world more like a human. Coupled with full-line chassis control (steering-by-wire, brake-by-wire, four-wheel steering, 800V active suspension), large models can directly control actuators without traditional MCUs.

“This system will be an important standard for the embodied intelligence of future vehicles,” Li Xiang said.

In 2025, Li Auto’s financial report was full of contradictions: revenue declined, profits halved, yet it maintained 100 billion yuan in cash and three years of profitability. But market patience is finite—falling out of the top three and halving profits mean no room for error.

A 20% growth target is both a bottom line and a gamble.

The new generation L9 not only represents an upgrade of the L series but also a test of whether Li Auto can prove that “software and hardware integrated” is more than just a slogan, relying on Mahe 100 and full-line chassis control. The pure electric i6 has broken through capacity bottlenecks, and orders for i8 are rising, but the real test will come in the second half—when the i9 enters the market, Li Auto will truly reveal its hand in the pure electric arena.

At that point, whether Li Auto’s “rebound” materializes will determine if it returns to the peak or is left behind by the wave of new competitors.

(Article by | Engine Perspective Author | Han Jingxian Editor | Li Yupeng)