Which Automaker Will Lead in 2026? We Break Down 13 Brands to Watch—and Those Falling Behind

In the year just ended—2025—China’s auto industry clearly turned a page. After several years of rapid expansion in technology and capacity, the competitive game is shifting from “spending hard and telling big stories” to “running efficiently and delivering real results.”

The old playbook that fueled the industry’s sprint is losing power fast. Relying only on lower prices, faster model launches, and louder marketing to win incremental growth is producing weaker and weaker returns.

What’s replacing it is a far more comprehensive contest: cash flow and profitability models, technology delivery and safety compliance, channel efficiency and organizational execution—and the ability to scale globally with real operations, not just ambition.

That’s why “looking ahead” can no longer be a routine year-end exercise. It has become something the industry must do earlier, more carefully, and with far more operational seriousness.

With that in mind, Huxiu Auto has outlined six key questions that are most worth judging in advance for China’s automotive sector in 2026. These questions also form a practical framework for observation—and under each dimension, different brands will emerge as “leaders” or “laggards,” depending on what exactly is being measured.

Market size: Can EVs and hybrids push past the 20 million threshold on stronger intrinsic demand? Where will the incremental engine come from—trade-in policies, lower-tier cities, PHEV/EREV growth, or faster charging and better energy replenishment? Which automakers are most likely to “break out” on sales and share, and which face the steepest uphill climb?

Profit expectations: Will the price war continue in another form? As automakers shift from chasing scale to chasing profit, which segment will become the fiercest battlefield? Which domestic incumbents and new entrants deserve the closest financial tracking?

L3 and intelligent driving: As L3 road trials begin, how will the supply chain and business models of intelligent driving change structurally? Are automakers truly prepared to carry clearer, more complete responsibility boundaries in “smart driving activation” scenarios? Which technical routes and supply-chain players matter most?

Luxury ambitions: In the EV era, how should “luxury” be rebuilt—through brand assets, product experience, service systems, or ecosystem capability? Which Chinese automakers have genuine system-level luxury competence rather than configuration stacking and price escalation?

Going global: After overseas plants and localization accelerate, can exports in 2026 achieve both volume growth and profitability improvement? Can Europe, Southeast Asia, and other markets replicate China’s former “China speed”? Which automakers have stronger overseas certainty and greater upside?

AI crossovers: With cross-industry moves intensifying, will 2026 produce new “variable-level opportunities” that reshape the competitive landscape?

Question 1: Domestic + overseas—can total volume break 20 million, and who captures the growth dividend?

On December 30, China’s NDRC and Ministry of Finance issued the 2026 policy notice for large-scale equipment upgrades and consumer trade-ins, formally landing the subsidy framework for automotive trade-ins.

That same day, the General Office of the Ministry of Commerce—together with seven other departments—released the detailed implementation rules for 2026 auto trade-in subsidies, specifying the mechanisms for both scrappage replacement and vehicle swap replacement.

For the auto industry, this is a tangible New Year’s gift. With expectations still cautious around consumer momentum, policymakers are not easing off the accelerator.

So what kind of growth can China’s NEV market see in 2026?

Cui Dongshu, Secretary-General of CPCA, told Huxiu that he expects around 10% growth for full-year domestic NEV sales—about 14.137 million units.

Zhang Yongwei, Secretary-General of the China EV100, is more optimistic. He believes China’s NEV sales in 2026 could reach 20 million units, including 3.5 million exports. That implies roughly 16.5 million NEVs sold domestically, a year-on-year increase as high as 28.4%.

Whether the true number is closer to 10% or 28%, one conclusion is hard to ignore: the biggest incremental dividend is likely to land in the RMB 150,000–200,000 price band.

Start with the policy signal. Compared with earlier policy rounds, the subsidy cap hasn’t changed—but the method has shifted from a fixed amount to a vehicle-price-based percentage.

Under the new structure, a consumer who wants to receive the full RMB 20,000 scrappage replacement subsidy would need to buy a vehicle priced at RMB 166,700 or above. To receive the full RMB 15,000 swap replacement subsidy, the purchase price would need to be RMB 187,500 or above. In practical terms, that means buyers in the RMB 150,000–200,000 bracket are positioned to receive the strongest effective policy boost.

Now look at the supply side. In 2025, “top-tier tech trickling down” became an industry-wide trend. Features that used to be reserved for models above RMB 200,000—400kW dual motors, lidar-equipped advanced driver assistance, Qualcomm 8295-class cockpit chips—were pushed down into RMB 150,000–200,000 products. That materially improves competitiveness in this band.

Demand is also large enough to support a real breakout. According to CAAM data, in the first 10 months of 2025, NEV sales in the RMB 150,000–200,000 price range reached 2.707 million units, ranking second among all price bands—only 32,000 units behind the top segment.

All of this makes RMB 150,000–200,000 the market’s “golden zone,” favored simultaneously by policy, supply, and demand.

On market share dynamics in 2026, here is how Huxiu views the landscape.

Leapmotor is already well-positioned in the RMB 150,000–200,000 band with three models: C11, C16, and C01. Both C11 and C16 have delivered solid market performance, meaning their competitiveness is already validated.

More importantly, from Q4 2025 into the first half of 2026, Leapmotor is expected to roll out additional new-series models, sustaining a high-frequency product offensive that increases saturation pressure on rivals. In 2025, Leapmotor is also highly likely to achieve full-year profitability—a rare feat among “new forces”—with both sales and profits rising. That matters, because profitability materially strengthens consumer confidence in brand durability.

For 2026, Leapmotor has set a bold target: one million units sold. Company executives have said that once they made this goal public, “there is no possibility of adjustment.”

GAC Aion, by contrast, is under obvious pressure. In the first 11 months of 2025, Aion’s cumulative sales reached 247,900 units, down 19.29% year-on-year. A slower response to market shifts has been a key factor behind the decline.

Since 2022, EREV and PHEV products have entered a rapid-growth lane and formed a crucial pillar of the NEV market. Yet Aion only launched its first EREV model—Aion i60—in November of last year. That delay meant that as Aion’s previous volume models entered a downward cycle, there was no new “next hit” ready to take over.

Compounding that is a brand problem Aion has not fully shaken off: the “ride-hailing car” label. Across auto media and social platforms, it’s easy to find content framing Aion as ride-hailing-friendly—and some dealers even promote it that way. This long-running narrative continues to hurt brand building and its performance among private consumers.

Tesla’s situation in China is also evolving. Its main-selling models have seen competitiveness weaken in the domestic market. In the first 11 months of 2025, Tesla sold 532,000 vehicles in China, down 7.4% year-on-year. Still, Tesla’s brand influence remains significant. And the rumored “Model Q” priced below RMB 200,000 remains one of the most consequential potential variables that could reshape competitive dynamics—if it becomes real.

Question 2: Will the price war ease—or simply change form?

Before December 12, 2025, when Huxiu Auto discussed 2026 expectations with industry insiders, the consensus was blunt: the price war would become even more intense.

That belief was built on three major factors.

First, purchase tax adjustments would raise the cost of buying: beginning this year, NEV purchase tax moved from “exempt” to “half-levied.”

Second, overall market growth was expected to slow. Zhang Yongwei, for example, forecast domestic auto market growth (ICE + NEV combined) at only 2%—far below the 9.7% year-on-year growth seen in the first 11 months of 2025.

Third, the 2025 price war was unusually fierce. In 2024, the average price of passenger cars in China was RMB 184,000. In the first 11 months of 2025, that average slipped to RMB 178,000.

Notably, despite intense price wars in 2023 and 2024, average passenger car prices still rose. But in 2025, the average fell by RMB 6,000—an unmistakable sign of how brutal the competition became.

Then, on December 12, a major variable emerged. China’s State Administration for Market Regulation released the “Compliance Guidelines for Price Behavior in the Automotive Industry,” aimed at regulating automotive pricing practices and explicitly prohibiting improper competition such as selling below cost.

After the draft was issued, more than a dozen automakers—including BYD, XPeng, BAIC, Changan, Great Wall, Chery, Dongfeng, Leapmotor, Seres, and JAC—publicly expressed support.

Cui Dongshu told Huxiu the draft guidelines could push automakers away from “bottomless price cutting” and toward competition strategies that prioritize profitability—helping improve industry-wide margins.

Skeptics argue that these are just statements, and price wars will continue anyway.

But even the market itself is sending a warning: discounts are delivering less and less demand stimulation. In November 2025, both NEV and ICE discount rates reached their yearly highs—yet passenger car sales fell 8.1% year-on-year and 1.1% month-on-month.

If price-war expectations soften, NEV makers are likely to raise the priority of profit, pushing more products above RMB 200,000. That would sharply intensify competition in the RMB 200,000–400,000 band—where Cui believes the fiercest battles may unfold in 2026.

Li Auto, NIO, and Xiaomi—brands focused on RMB 200,000+—all plan new product launches this year. Even Leapmotor, known for value-for-money positioning, is preparing to launch its highest-end D-series models.

A softer price war does not mean automakers will stop cutting costs. For weaker suppliers, the outlook could worsen. Some suppliers worry that automakers, to avoid crossing the “below production cost” red line, may shift cost pressure downstream by further squeezing parts procurement prices—preserving terminal pricing room while intensifying survival pressure on vulnerable suppliers.

On the profit outlook for 2026, Huxiu offers the following views.

Geely is one of the brands most worth watching. In 2025, its sales surged past 3 million units. Based on disclosed data from the first three quarters, even as volumes climbed, Geely’s gross profit and average selling price remained stable—or even improved.

In 2026, Geely’s profit growth could come from four sources.

First, post-privatization consolidation of Zeekr could directly lift net profit attributable to shareholders, while the “One Geely” system accelerates integration across four major brands in R&D, procurement, and channels—driving clear cost reductions and efficiency gains.

Second, volume expansion for new models under Galaxy and Zeekr could lift both sales and ASP. With 2026 sales potentially targeting 3.45 million units, a higher mix of mid-to-high-end models could raise net margins.

Third, after the integration of Qianli Technology, internal resource friction may ease. Platform-based manufacturing and scaled procurement could further dilute costs—though the actual integration results remain to be seen.

Fourth, strong overseas growth and asset optimization moves such as share buybacks could strengthen profit stability. Minsheng Securities has forecast Geely’s revenue at RMB 404.78 billion, 489.69 billion, and 572.83 billion for 2025–2027, with net profit attributable to shareholders at RMB 16.21 billion, 22.09 billion, and 25.97 billion.

Geely also faces multiple challenges: ongoing price and product pressure from BYD and Huawei-affiliated competitors; macroeconomic uncertainty, trade policy changes, and FX volatility impacting overseas performance; and continued strategic trade-offs between BEV, PHEV, and other technology routes, where the precision of bets will shape long-term profitability.

For GAC, 2025 was far heavier. Sales and financial performance both struggled. In the first three quarters, GAC’s revenue fell 10.49% year-on-year to RMB 66.272 billion, while net loss reached RMB 4.312 billion—versus a RMB 120 million net profit in the same period a year earlier.

Organizationally, from September onward, reports of senior executives being summoned for talks, taken away for investigation, or resigning abruptly became frequent.

GAC’s leadership is not standing still. At its 2026 work conference in December, Chairman Feng Xingya delivered a keynote titled “Break Through via Transformation, Unite to Win, Strive to Rebuild a New GAC.” He positioned the meeting not only as a deployment session, but also as a mobilization for confronting challenges and attacking transformation. His message emphasized strategic focus, internal courage for reform, unified execution, and a decisive posture for breakthroughs.

For an enterprise of GAC’s scale, transformation is typically a long, difficult marathon. In 2026, many key indicators may still look unfavorable.

Xiaomi Auto sits at the opposite extreme.

Since launching its first model, the Xiaomi SU7, in April 2024, Xiaomi’s auto business has reported notably strong financial performance. In the first three quarters of this year, Xiaomi’s overall financial growth remained rapid, with no visible inflection point downward yet.

The concern now is whether new orders are declining quickly. Recent signals from multiple sources suggest Xiaomi’s weekly orders have dropped to around 4,000–5,000 units, while its monthly deliveries have climbed to 50,000 units—and could rise to 70,000 units in early 2026. The gap between deliveries and new orders would have to be filled by the previously accumulated backlog.

That raises the possibility that Xiaomi could shift from a “supply can’t meet demand” phase to a “demand-driven” phase where sustaining momentum becomes the real test.

Question 3: Once L3 hits public roads, how will the intelligent-driving ecosystem change?

On December 15, 2025, China announced its first batch of L3 “conditional automation” model approvals. Two models—one from BAIC and one from Changan—were permitted to conduct road trials in designated areas in Beijing and Chongqing, targeting highway segments and urban congestion scenarios respectively.

The most important meaning of L3 is not technical—it is legal and structural: responsibility begins to shift from the driver to the automaker. When the L3 system is operating, liability is borne by the automaker; when the driver takes over, liability returns to the driver.

In 2026, will China see its first case where an automaker is penalized for responsibility after an L3-related accident? It’s unclear. But when those cases inevitably occur, the controversies they trigger will matter deeply for how society enters the true smart-driving—and eventually autonomous-driving—era.

Huxiu’s key judgments are as follows.

First, L3 will drive systemic transformation across the entire automotive ecosystem—vehicle design, components, software algorithms, communications, and data services. The ripple effects will reshape vehicle architecture, supply chains, and production models.

At the vehicle design level, dual-redundancy architectures for critical execution systems such as braking and steering are likely to become standard.

At the supply-chain level, core intelligent components such as lidar and high-compute chips will become rigid requirements. Cross-domain integration will also push Tier 1 suppliers beyond single-domain delivery and toward multi-domain, full-stack solution partnership roles.

At the production-model level, “hardware pre-embedding + software iteration” will become mainstream—forcing automakers to build full-chain capabilities spanning R&D, testing, and operations, and accelerating the industry’s shift from “manufacturing-oriented” to “ecosystem service-oriented.”

Second, L2 will see a new wave of “standardization,” while L4 begins to break the ice.

In 2025, L2 functionality “democratization” started strong but faded under tighter compliance pressure around marketing and road legality. With L3 arriving, however, L2 features are likely to become standard equipment for NEVs.

That standardization wave will also move more forcefully into the ICE market. In 2025, brands such as SAIC Volkswagen and Great Wall’s Haval had already started promoting “smart driving for both ICE and EV.” By 2026, that trend—“ICE cars can be smart too”—will likely intensify as L2 becomes standard.

Huawei’s Intelligent Automotive Solution BU CEO Jin Yuzhi has estimated that by 2027, penetration of assisted-driving models supporting highway navigation with lane changes plus urban capability could exceed 50%, while highway L3 penetration could surpass 20%.

In other words, in 2026, ICE vehicles with L2 will become meaningfully more competitive—and NEVs without L2 will look increasingly awkward.

Many in the industry also argue that L2, L3, and L4 are not a simple linear ladder. As L2 scales into mass adoption, higher-level L4 could also advance quickly in parallel.

Horizon founder and CEO Yu Kai has predicted L4 could appear within three to five years. XPeng Chairman He Xiaopeng holds a similar view, arguing the future is L2 and L4 rather than L3, because “L3 is neither L2 nor L4.” He even publicly “bet” with his team that by August 30, 2026, XPeng’s VLA should reach the overall effect of Tesla’s FSD 14.2 in Silicon Valley.

In 2026, China’s intelligent-driving rollout, trickle-down, and competitive comparison may become a “many flowers bloom” landscape.

For 2026 smart driving, Huxiu’s view is this.

Even before 2026 begins, Huawei Qiankun has already launched L3 road testing on real roads in cities such as Chongqing and Hefei, covering both urban and highway scenarios. In licensing, pilot implementation, and solution deployment, Huawei Qiankun clearly holds the strongest first-mover advantage.

By official figures, 35 models already carry Huawei Qiankun’s intelligent-driving system, spanning a wide RMB 160,000 to RMB 1 million price range.

In 2026, China’s L3 landscape may well become “Huawei and everyone else.”

If Huawei represents the “Mount Everest” of performance, then Zhoyu may represent the “sea level” of scale and accessibility. By the end of 2025, Zhoyu’s cooperating models reportedly exceeded 80. FAW has also become Zhoyu’s largest single shareholder. Recently, Zhoyu signaled ambitions to accelerate into heavy-truck NOA—an aggressive expansion path.

Two major players also deserve attention because they face particularly high pressure.

In China’s urban NOA market, a multi-structure competition has formed: in-house automaker development, top-tier third-party platforms, and diversified suppliers coexisting.

Geely’s pressure lies in quickly outputting consistent, stable intelligent-driving experiences after centralizing integration under Qianli Technology—so it can establish clearer technical labeling and user reputation.

BYD’s pressure is rooted in the bottleneck between having a data advantage and converting it into a true experience advantage. Even if BYD generates 150 million kilometers of assisted-driving data per day, extracting valuable long-tail scenario information from that ocean of data remains a core challenge. That places extremely high demands on its data middle-platform processing capability.

Compared with leaders like Huawei, BYD’s performance in complex scenarios—put politely—still has “significant room for improvement.”

As L3 deployment accelerates in 2026, BYD will also need to demonstrate stronger capabilities in defining human-machine takeover boundaries and building safety redundancy systems—critical indicators of its intelligent-driving maturity.

Question 4: How will China’s luxury narrative be told next?

Here, “luxury” is discussed specifically within the NEV space.

How do you define luxury? Under common industry association standards, models priced above RMB 300,000 are categorized as luxury. But it’s also important to acknowledge that the fuel era’s BBA-drawn brand boundary has been blurred by price war smoke. Rapid iteration in smart driving and smart cockpits has made the definition of NEV luxury even more diverse and complex.

As L3 conditional automation pushes forward, the most direct impact is that L2-level capabilities will trickle down faster and more broadly.

So when a RMB 100,000 car and a RMB 500,000 car differ by only 5% in smart-driving experience, how can Chinese NEV luxury products—once relying on “experience premium” and “compute premium” from intelligent features—tell a new story in 2026?

This is a difficult question for Chinese brands that have not yet truly mastered the “luxury path” or told a convincing luxury story.

Huxiu’s first judgment: “pseudo-luxury” will have nowhere to hide in 2026.

In recent years, Chinese NEV brands have tried to climb upward, launching everything from RMB 400,000–500,000 luxury models to million-RMB ultra-luxury products. Most have proven short-lived. Some models now sell fewer than 500 units per month.

Great Wall Chairman Wei Jianjun has argued that China still has no true high-end auto brand—only high-end products. In his view, a premium brand requires time and cultural accumulation, anchored by spiritual-level value propositions recognized by users.

These propositions must be built through product quality, service quality, and user trust.

In reality, while China has launched many strong single high-end models, brands often struggle to maintain stable performance in ongoing software upgrades, channel building, and service capability. As a result, many once-ambitious luxury NEVs look exhausted by late 2025, with their “high-end road” increasingly uncertain. As industry insiders put it: “Anyone can stack specs; the real skill is selling well and sustaining it.”

In 2026, more pseudo-luxury products may be forced off their pedestal.

Second judgment: tech companies will be more cautious about pushing high-end products.

Technology is making “luxury” harder to sell.

Leapmotor founder Zhu Jiangming recently reiterated that luxury narratives will only become harder under the trend of smart-driving democratization—and that the “3C-ization” of cars is unstoppable. In his view, there is no luxury in 3C products, only mass-market categories—and cars will move the same way. That is why Leapmotor is firmly not taking the “luxury route.”

This isn’t just an opinion; the data supports it.

CPCA figures show NEV penetration in the RMB 100,000–200,000 band rose from 11.9% in 2022 to 19.7% in 2025. The RMB 200,000–300,000 band rose from 4.6% to 10.8%. But above RMB 300,000, growth is far weaker: RMB 300,000–400,000 rose from 1.1% to 3.4%, and above RMB 400,000 rose from 0.9% to 1.3%.

In other words, growth in penetration above RMB 300,000 is modest. By comparison, NEV penetration below RMB 100,000 rose from 9% in 2022 to 18.1% in 2025—the largest increase across all tiers.

This reflects two realities: consumption downshifting is visible, and NEV makers are attacking the ICE market through “dimensionality reduction,” squeezing fuel competitors with strong products at lower prices. That strategy has also pushed players to concentrate intensely under RMB 200,000—turning the red ocean into a blood ocean.

In 2026, under pressure from profitability and competition, market share and cash flow will become top priorities for tech-oriented automakers. Limited production resources will be allocated toward high-volume products, making high-end launches more cautious and more pragmatic.

If technology is dulling luxury’s shine, then “high-end manufacturing” will become the sturdier foundation for luxury storytelling—Huxiu’s third judgment.

Once smart driving becomes the new era’s “automatic transmission,” the core contest still returns to fundamentals: stamping, welding, painting, and assembly. A premium car ultimately needs more precise and intelligent manufacturing, and stronger supply-chain capability.

JAC has a strong example here.

The Zunjie “super factory” reportedly involved over RMB 10 billion in investment and operates as a full-link 5G factory. It integrates Huawei’s iDME industrial digital model-driven engine technology, with high automation across stamping, transfer, and packaging systems. More than 1,800 intelligent robots enable highly automated and flexible customization across the four major processes. Its digital twin system captures 300,000 data points per second and builds a full lifecycle “electronic file” for each vehicle.

With Huawei’s halo plus strong intelligent manufacturing, the ultra-luxury Zunjie S800—priced from RMB 708,000 to RMB 1,018,000—has reportedly surpassed 10,000 cumulative deliveries since its launch on May 30, 2025, and is aiming for monthly sales above 4,000 units.

By market performance, the Zunjie S800 is currently one of the most successful Chinese luxury cars.

For 2026 luxury narratives, Huxiu’s view is this.

Huawei’s deep full-stack empowerment, plus the proven path Zunjie has built across R&D, manufacturing, delivery, and after-sales service, gives its flagship MPV a stronger chance to further disrupt the million-RMB luxury segment in 2026.

Avatr, however, is likely to face heavy pressure in the luxury market in 2026.

In sales terms, Avatr sold 128,800 units in 2025, achieving only 58.5% of its annual target (220,000). Higher-end models Avatr 11 and 12 underperformed, while volume largely depended on Avatr 07—insufficient to strongly support a luxury positioning. In brand perception, the “Changan + Huawei + CATL” alliance has not solidified into a clear consumer cognition; the blurred positioning weakens momentum.

That said, as a high-end NEV brand under Changan with strong state-backed resources, Avatr has real support in intelligent manufacturing, supply chain, capital, and technological synergy.

The breakthrough path in 2026 depends on stepping away from self-congratulation and finding a sharply defined target segment—translating Huawei smart driving and CATL battery tech into clear scenario value, building a crisp brand label, and executing with both state-owned strength and market mechanisms. If it can do that, it still has a chance to stabilize its footing in the luxury lane.

Question 5: Going global is the obvious direction—but who can truly win?

The latest CPCA data shows that from January to November 2025, China’s auto exports surpassed 7.33 million units, with passenger vehicle exports reaching 6.23 million units. Conservatively, full-year figures are likely to exceed 8 million total exports and 7 million passenger exports—up around 25% and 20% year-on-year respectively.

Compared with 2024, sales increased across almost all export destinations. Growth was particularly strong in Mexico (a major distribution hub for Latin America) and the UAE (a key hub for consumption and distribution across the Middle East), both large in volume and fast in growth. Markets such as Brazil, the UK, Australia, Saudi Arabia, and the Philippines also maintained steady increases.

At the same time, new energy penetration in exports reached 53.6% in the first 11 months of 2025—up roughly 30% year-on-year—highlighting China’s product strength in NEVs.

2025 was also a year when Chinese automakers’ overseas capacity began to translate into real output.

Various incomplete estimates suggest that overseas plants entering production this year alone represent more than 1.2 million units of planned capacity. The driver is straightforward: Chinese cars are facing supply gaps in overseas markets. Building capacity abroad not only serves demand better in key markets, but also helps mitigate trade barrier impacts.

However, overseas capacity is only one condition for scaling exports.

To truly globalize brand and volume—looking at Japanese and Korean giants from the fuel era—automakers must build an integrated system: overseas manufacturing bases, shipping fleets, and global dealer networks.

Toyota, still the global volume leader, has more than 75 production bases across five continents. Beyond factories, it relies on over 60 roll-on/roll-off ships operated by companies such as Toyota Tsusho, Nippon Yusen, and Kawasaki Kisen for global logistics. It also maintains a complex dealership network spanning 190 countries and regions.

BYD’s Asia-Pacific Auto Sales GM Liu Xueliang previously told Huxiu that the auto industry ultimately depends on long-term accumulation and repeated trial-and-error to form a complete system. Toyota isn’t strong because it can produce 10 million cars—it’s strong because every surrounding link can support 10 million cars.

Today, the impression Chinese cars leave on overseas consumers is still largely centered on “good value,” “more choices,” and “lower prices”—important traits, but still relatively basic.

Even Liu estimates that China’s true globalization will take multiple “generations” of industry iteration, with each generation spanning roughly 10–20 years.

For 2026 NEV globalization, Huxiu’s view is this.

BYD, through building plants in Hungary and Turkey, is poised to shift from an “import supplier” to a “local European manufacturer,” potentially avoiding a 27% tariff burden while gaining larger strategic value. The two sites form a north-south European layout that also diversifies risk. Once the Hungary and Turkey plants enter production—alongside existing capacity in Thailand, Brazil, and elsewhere—BYD’s exports in 2026 could see a step-change leap.

Question 6: Is crossing into AI a must-win battlefield?

In Q4 2025, Li Auto released its AI smart glasses, Livis. Internally, Li Auto placed it under a newly formed “wearable robotics” department established in June. Officially, Livis is described as a “wearable robot on your head.” It starts with a display-less form to optimize weight and battery life, and plans future development with Zeiss for a display version—moving toward an “independent terminal that doesn’t need a phone.”

For now, though, the product looks more like an extension of in-car control and infotainment rather than a truly independent wearable robot. The launch highlighted solving the AI glasses “impossible triangle” of weight, battery, and performance, but practical functionality still relies heavily on Li Auto’s ecosystem—limiting appeal for non–Li Auto owners.

More fundamentally, there is no decisive reason why automakers must lead this hardware category. Smartphone manufacturers and optical specialists can follow quickly; automakers’ first-mover advantage is mostly limited to car-connection scenarios.

As for older crossover routes like aircraft and embodied intelligence, the distance to full commercialization is still significant.

At XPeng AeroHT, the “Land Aircraft Carrier” modular flying car factory has been topped out and aims for mass production delivery in 2026. It reportedly has 260 orders, offers over 1,000 km range for the ground module, and uses electric VTOL for the flight module. Yet the broader industry outlook suggests that even by 2025 it has only entered “commercialization 1.0.” Large-scale eVTOL applications may not arrive until around 2035, and deep integration of low-altitude transport with ground mobility could take until around 2050. The actual landing pace is far slower than popular imagination.

XPeng’s IRON humanoid robot also drew attention after debuting in October—featuring a 22-DoF dexterous hand, 82 joints, and 2,250 TOPS compute, with capabilities like stylized walking and demonstrations intended to prove it’s not a human in costume. The goal is mass production by the end of 2026. But even XPeng’s robotics leadership has acknowledged that some of the current performance is an emergence of combined hardware and software—and that reproducing such emergence reliably is not yet fully understood.

Compared with these, Tesla’s Grok-based conversational capabilities may be more immediately useful to consumers. In an OTA update in early December, Tesla expanded Grok’s permissions so it could do more than “chat” in the car—it could add or edit navigation destinations, evolving into a more practical “personal travel assistant.”

This approach—using AI to deeply optimize core driving and travel scenarios to raise service capability—is likely the direction Chinese automakers should study in 2026, rather than pursuing flashy and premature crossovers.

In Huxiu’s view, AI crossovers will not become automakers’ primary strategic focus in 2026, and the odds are low that dramatically more impactful “new crossover variants” will emerge.

Simply put: in the competitive reality of today’s auto market, it isn’t the most important battle.