WeChat, Xiaohongshu, and Lili are all starting to make money. Where should they go from here?

By | Vision New Research Institute Wei Qiyang

Xiaopeng, NIO, and Li Auto have all started making money.

Not long ago, with Xiaopeng Motors’ top-line release of its 2025 financial report, China’s new-energy vehicle makers entered a historic moment—NIO, Xiaopeng, and Li Auto all achieved “full-company profitability” in their financial reports for the first time.

Although the “profitability” of the three companies varies in value, it marks the end of the rough growth phase in which new-energy makers relied purely on financing and burning cash, and the start of a new cycle focused on refined operations and self-sustaining cash generation.

At this key turning point from “scale expansion” to “profit orientation,” the three leading companies once collectively called “Xiaopeng, NIO, and Li Auto” actually followed distinctly different profit trajectories. Once you peel back the packaging of the financial reports, you see three divergent business paths—and the future divergence they foreshadow.

01 Full-company profitability: worlds of ice and fire

In 2025, NIO, Xiaopeng, and Li Auto achieved “full-company profitability” in their financial reports for the first time. But if you dig into the individual financial figures, the three companies’ fundamentals show a clear step-like pattern—each has highlights, and each also has hidden worries.

Core data comparison of Li Auto, NIO, and Xiaopeng financial reports Table: Vision New Research Institute

Li Auto remains the player with the largest revenue scale. In 2025, Li Auto generated revenue of RMB 112.3 billion, crossing the RMB 406.3k threshold for the third consecutive year and continuing to lead the revenue ranking among new-energy players.

However, beneath the surface of “defending the throne,” there is a current of under-the-radar turmoil.

In terms of deliveries, Li Auto delivered only 406.3k vehicles for the full year, down 18.8% year over year—becoming the only company among the three leading new-energy players with negative sales growth. This figure implies that Li Auto’s revenue growth did not come from selling more vehicles; instead, it relied on improving the mix toward higher-priced models to barely support the numbers.

What worries the market most is its profitability. Full-year net profit was only RMB 1.1 billion, plunging by more than 85% year over year. If you look at it from the perspective of “net profit margin under 1%,” Li Auto in 2025 was essentially walking along the break-even line.

Even more worth examining is what makes up that RMB 1.1 billion in profit—of which interest and investment gains totaled as much as RMB 1.92 billion. In other words, if you strip out this non-operating portion, Li Auto’s core business in 2025 is effectively approaching the edge of losses. That means Li Auto—once known as the “new-energy player that makes the most money”—turned in a “profited but nothing to show” performance in 2025.

NIO ended a loss “marathon” lasting 11 years, but its full-year figures are still not encouraging. In 2025, NIO generated revenue of RMB 87.49 billion, up 33.1% year over year, and narrowed its net loss to RMB 14.94 billion. From a full-year perspective, NIO remains the company with the largest loss among the three.

The real highlight came in the fourth quarter. Revenue that quarter was RMB 34.65 billion, a record high; operating profit reached RMB 1.25 billion, marking the company’s first single-quarter profitability.

Behind this “miracle reversal” was the concentrated surge of high-end models. The all-new ES8 became the profit driver in the fourth quarter, with single-quarter deliveries of about 39.7k units, accounting for 31.8% of total deliveries. NIO officially disclosed that the ES8’s whole-vehicle gross margin exceeded 25%, far above the company’s overall level. It was this flagship SUV priced above RMB 500k that pulled NIO’s whole-vehicle gross margin in the fourth quarter up to 18.1% on its own.

However, NIO’s hidden concerns are just as evident. A loss of RMB 14.94 billion for the full year means the cash burn in the first three quarters was still severe. Whether fourth-quarter profitability can be sustained remains unknown.

In addition, NIO’s large battery-swap network and offline service system are still consuming substantial capital. By year-end, NIO had more than 2,400 battery-swap stations and nearly 500 offline stores. Whether these heavy assets can truly translate into profit will be a key question for 2026.

Li Auto, NIO, and Xiaopeng sales data comparison Table: Vision New Research Institute

Xiaopeng was the growth champion in 2025 deliveries. For the full year, Xiaopeng delivered 429.4k vehicles, up 125.9%, generating revenue of RMB 76.72 billion, up 87.7%. Although it still posted a net loss of RMB 1.14 billion, the year-over-year narrowing was significant at 80.3%, putting it just one step away from turning full-year profitable.

Xiaopeng’s fourth quarter was also strong. It achieved single-quarter profitability of RMB 380 million for the first time, with a gross margin of 21.3%, ranking first among the three companies’ quarterly gross margins.

Behind this result were two legs of the approach: first, volume from an affordable blockbuster, the MONA M03. In the fourth quarter, this model delivered more than 60k vehicles, accounting for over 45% of Xiaopeng’s total deliveries that quarter, successfully diluting fixed costs; second, technology monetization has begun to pay off. In 2025, electronic and electrical architecture technology services that Xiaopeng provides to Volkswagen contributed more than RMB 3.0 billion in high-gross-margin revenue.

Overall, even though “full-company profitability” sounds impressive, the underlying bases of the three companies’ financial data are clearly different. Li Auto is the “defender”—profitable, but with a shaky foundation; NIO is the “comeback maker,” completing a dramatic reversal by year-end through blockbuster products; while Xiaopeng is the “pursuer,” quickly closing the gap through economies of scale and technology monetization.

02 Three completely different “money-making” logics

Judging from market performance, the paths these three companies took to achieve profitability, to a certain extent, reflect the divergence of their underlying business logic.

Li Auto’s profits dropped sharply mainly because the “product definition” capability that previously enabled results has been replicated by competitors. Put more directly, the market has become more crowded, and its competitive advantages have been weakened.

Before 2023, Li Auto’s L series relied on a precise positioning—“a six-seat SUV with no range anxiety”—and had almost no competitors in its submarket. But the competitive landscape in 2025 is no longer the same. Models like Aito M7 and M9 directly engaged Li Auto in the RMB 300k–500k price band by leveraging Huawei’s intelligent driving and channel advantages. Aito M9 delivered more than 150k units in 2025, directly eroding Li Auto L9’s market share; meanwhile, Leapmotor’s C16, priced at around RMB 150k, also pulled six-seat extended-range SUVs down into the lower-tier market.

The diversion of market share is reflected directly in the financial data. Li Auto L8 terminal discounts expanded from RMB 5,000 at the start of the year to RMB 25k by the fourth quarter. L9 likewise increased promotional intensity: the average selling price per car fell from about RMB 278k in the third quarter to around RMB 250k in the fourth quarter. Vehicle gross margin then dropped below the healthy benchmark of 20%, falling to 16.8%.

As for the pure-electric strategy that had been heavily expected, the launch faced an unfavorable start. Li Auto’s first pure-electric product, MEGA, underperformed expectations due to a public opinion controversy. The pure-electric SUV i8 accumulated fewer than 15k deliveries in the first three months after launch, and the i6 was forced to postpone deliveries because of battery supply bottlenecks.

Li Auto’s 2025, in essence, was the pain period of “the old engine stalled, the new engine not yet fired.”

Next, let’s look at NIO. Its profitability path comes from boosting high-end blockbuster models and implementing cost reduction and efficiency gains.

First is optimization of the product mix. In the fourth quarter of 2025, NIO decisively tilted resources toward high-gross-margin models. The all-new ES8 delivered about 39.7k units in the single quarter, accounting for 31.8% of total deliveries. This flagship SUV, with a starting price of RMB 498k, has a whole-vehicle gross margin close to 25%, becoming NIO’s “profit cash cow” in the fourth quarter. Low-gross-margin models such as the ET5 had their production schedule compressed to a certain extent; this strategy directly lifted the overall gross margin.

Second is the full implementation of cost reduction and efficiency improvements. In the fourth quarter, NIO carried out strict expense controls: research and development expenses fell sharply by 44.3% year over year, and sales and administrative expenses decreased by 27.5% year over year. According to NIO insiders, in the fourth quarter the company also reduced period expenses to near the lowest level in nearly two years through measures such as freezing part of recruitment and cutting travel budgets.

At the same time, cost reduction through technology has also started to pay off. The self-developed “Shenji NX9031” autonomous driving chip was officially put into use in the fourth quarter. This 5nm chip replaced the previously needed Nvidia Orin-X solution that required 4 chips per vehicle, bringing about RMB 10k of cost optimization per car. Using the fourth quarter delivery volume of 125k vehicles, this single chip self-development alone saved NIO more than RMB 1.2 billion in costs.

Different from Li Auto and NIO—indeed, very different from the traditional auto-industry profitability model—even the way Xiaopeng profits is to “sell technology” to supplement “selling cars.”

In 2025, Xiaopeng’s service and other income reached RMB 8.34 billion, up 65.6% year over year, and its share of total revenue first exceeded 10%. This portion of revenue mainly comes from three parts: electronic and electrical architecture technology services provided to Volkswagen—about RMB 3.2 billion for the full year; parts sales—about RMB 2.8 billion; and carbon credit trading—about RMB 2.3 billion.

In its cooperation with Volkswagen, Xiaopeng supplies Volkswagen’s CEA electronic and electrical architecture to the Hefei plant. The per-vehicle technology service fee is about RMB 4,000. Based on Volkswagen’s plan for annual production of 300k units, this single item can generate technology revenue of RMB 1.2 billion per year for Xiaopeng. The gross margin of the technology authorization business is as high as over 60%, greatly improving Xiaopeng’s financial structure.

On the vehicle side, Xiaopeng completely reversed its sales downturn with the affordable blockbuster MONA M03. In the fourth quarter of 2025, MONA M03 delivered more than 60k units, accounting for over 45% of Xiaopeng’s total deliveries that quarter. This compact sedan with a starting price of RMB 119.8k lowered Xiaopeng’s average selling price per vehicle to RMB 164k. But through scale dilution of R&D and manufacturing costs, the manufacturing cost per vehicle fell 18% year over year.

Three profit paths, three business logics.

Li Auto’s dilemma lies in its “defender” mindset failing to respond to changes in the competitive landscape; NIO’s breakthrough relies on concentrating on high-gross-margin blockbuster products, leaving no slack in cost controls, and timely realization of technology-driven cost reduction; Xiaopeng’s comeback is a kind of “misaligned competition,” opening a second growth curve through technology output, using the scale of affordable cars to support technology investment.

03 Three distinct routes for breakthrough

Because the logic behind “making money” differs, Xiaopeng, NIO, and Li Auto’s subsequent development strategies also show clear divergence. This is not only each company’s path choice, but also reflects their different understanding of the endgame of the auto industry.

Comparison of Xiaopeng, NIO, and Li Auto vehicle strategies Table: Vision New Research Institute

Xiaopeng is betting the future on the full rollout of “physical AI.” In 2026, Xiaopeng plans to increase AI-related R&D spending to RMB 7.0 billion, accounting for more than 80% of its full-year R&D budget. In He Xiaopeng’s plan, Xiaopeng is no longer simply an automaker—it is an “AI-driven robotics company.”

In terms of concrete deployment, Xiaopeng will introduce four new models within the year, covering the RMB 100k to RMB 400k price range; it will conduct Robotaxi passenger-carrying pilot operations in cities such as Guangzhou and Shenzhen; and it will push forward mass production of the humanoid robot IRON.

Supporting all of this is its self-developed Turing chip. The chip achieved mass shipments in the fourth quarter of 2025, with compute power of 750 TOPS per chip. The second-generation VLA large model is accelerating rollout, targeting an end-to-end closed loop from “perception” to “control.”

This full-stack layout gives Xiaopeng very strong “tech-stock” narrative capability in the capital market. But the risks are equally obvious: in 2025, Xiaopeng’s automotive sales gross margin was only 13%, far below Li Auto’s 16.8% and NIO’s 18.1%. With high-intensity R&D spending layered on low-gross-margin volume models, whether it can be sustained in financial terms is Xiaopeng’s ultimate test.

NIO’s strategic keywords for 2026 are “capacity expansion” and “battery-swap profitability.” Unlike Xiaopeng, which pursues technology spillover, NIO chooses to keep digging its moat deeper in the high-end market.

On the product side, NIO plans to launch three large vehicles: the flagship SUV ES9, a luxury sedan ET9 targeting the Mercedes S-Class, and a newly facelifted ES6. Internally, NIO calls this lineup the “three high-end swordsmen,” targeting a total addressable market size 2 to 3 times larger than in 2025. At the earnings call, Li Bin revealed that the ES9 is equipped with a 150-degree semi-solid-state battery pack, extending NEDC range to over 1,000 kilometers.

On the technology side, the second-generation “Shenji” chip has successfully taped out. Using a 5nm process, one chip can save NIO about RMB 8,000 in Nvidia Orin procurement costs.

On the energy replenishment side, NIO plans in 2026 to add 1,000 battery-swap stations, bringing the total number to more than 3,500. It will also push the battery-swap business to transform from a “cost center” to a “profit center”—by charging service fees to cover operating costs, and simultaneously outputting solutions to battery-swap alliance partners to earn technology authorization fees.

NIO’s advantage lies in its high-end brand mindshare and user loyalty. By the end of 2025, the user repurchase rate reached as high as 42%. But its disadvantages are also clear: overall scale is still unlikely to break through the million-vehicle threshold, and in the context of “ultra-fast charging” becoming widespread, the battery-swap model is facing efficiency challenges. When an 800V ultra-fast charge can replenish 500 kilometers of range within 15 minutes, the advantage of swapping in 5 minutes is being rapidly narrowed.

Li Auto, meanwhile, is trying to force a switch from “product-driven” to “technology-driven.” In 2025, its R&D spending reached RMB 11.3 billion, a record high, with AI-related spending accounting for more than half—so much so that this figure even exceeds the concurrent R&D spending of Xiaopeng and NIO.

In 2026, Li Auto will launch the facelifted L9 and the high-end pure-electric i9. Internally, the i9 is viewed as a “correction for MEGA.” Compared with MEGA’s forward-looking but controversial exterior, the i9 returns to the stable design language of the L series. It is equipped with its self-developed “Mach 100” chip and the VLA large model.

Li Auto’s biggest backing is its cash reserves exceeding RMB 39.7k (RMB 101.2 billion). This is its largest capital for enduring the pain period of the pure-electric transition.

04 Conclusion

For “Xiaopeng, NIO, and Li Auto,” the profitability shown in 2025’s financial reports is not the end of victory—it is the starting gun for a new round of elimination.

Xiaopeng is trying to transform from an automaker into a technology giant by investing heavily in AI technology and building high barriers. It is a path with high returns but high risk;

NIO is in the high-end market and in the moat of its battery-swap network, working to find the optimal solution for scale and profit. It is a steady path, but with an obvious ceiling;

Li Auto is doing everything it can to implant technology into its DNA to make up for the gap left after its product-definition approach was imitated. It is a compelled transition path, but with the richest resources.

The once stable three-way duopoly of “Xiaopeng, NIO, and Li Auto” is being thoroughly broken. In the endgame race to define the auto industry with AI, whoever can first cut through the fog of profitability and build sustainable dual-wheel drive of both technology and business will truly remain at this table.

Vast amounts of information and precise interpretation—everything is in the Sina Finance APP