Chinese car companies flex muscles at CES

```

Author | Zhou Zhiyu

Editor | Zhang Xiaoling

Leveraging the stage of CES, Chinese automakers are redefining the future.

Compared to previous years' CES (International Consumer Electronics Show), this year Chinese auto companies and supply chain enterprises appeared less noisy as newcomers and more pragmatic. If two years ago the Chinese auto industry was still trying to prove its "cost-performance" to the world through battery cost and supply chain efficiency, then at this year’s show, the core topic has completely shifted.

From on-site signals, it seems that simple hardware stacking has reached the threshold of the law of diminishing returns.

When the number of screens and 0-100km acceleration can no longer excite the capital market, Chinese automakers have collectively begun a charge towards the top of the industry chain—AI large models and advanced intelligent driving architectures. At this year’s CES, Geely released its "All-domain AI 2.0" system; Great Wall Motors brought intelligence and power; supply chain companies broke through in chip and sensor performance—together forming the Chinese narrative of CES 2024.

These core players from China are building new technological moats through advanced intelligent driving and AI large models. This also signals the thrilling leap in the restructuring of the auto industry: Chinese players moving from manufacturing output to standard output.

In Las Vegas, a new battle for value distribution rights in the automotive industry has quietly begun. As hardware homogenization becomes industry consensus, Chinese automakers are pouring investment into “soft power.” This is not just a product-level arms race; it’s also a deep overhaul in supply chain integration and global strategy.

Technological Challenge

At CES, it was evident that Chinese automakers and supply chain companies are launching a “saturation attack”—not only strategic pressure from industry giants but also focused breakthroughs from niche champions.

Geely Holding Group’s actions at CES mirror its intent for deep integration of R&D resources. Geely did not simply release a new car but launched a technological foundation: All-domain AI 2.0, centered on WAM (World Action Model).

Unlike end-to-end models, Geely’s WAM goes a step further in logic. Li Chuanhai, Vice President and President of Geely Automobile Research Institute, noted the industry is basically unified on the direction of end-to-end large models, including VLA (Vision-Language-Action). But WAM adopts a layered design, enabling a complete intelligent closed-loop from “understanding-planning” to “rehearsal-judgment-correction.” He emphasized that this gives automotive intelligence its first “worldview” and “judgment” for ongoing evolution.

As the “One Geely” strategy continues, innovation in technology can be more efficiently applied to products. Geely Auto Group CEO Gan Jiayue revealed that through this round of integration, Geely reduced the number of projects in product planning by about 20%–30%, while improving quality.

This “subtraction” logic aims to solve the internal friction problem of a multi-brand strategy. Gan Jiayue made it clear that although Zeekr, Lynk & Co, and Geely Galaxy target different markets, they now share the unified “Qianli Haohan G-ASD” smart driving software. On the hardware side, various tiers from H1 to H9 remain, but the software kernel is unified. This means new models like Zeekr 9X and Galaxy M9 will share the same AI brain.

“Some good technologies can empower more products. For example, Zeekr is currently using the ShenDun Gold Brick battery, which can later be used in Lynk & Co and Galaxy products,” Gan Jiayue explained. This tech-backfeeding mechanism enables models like Lynk & Co 900 to first feature Nvidia Thor-U chips, quickly channeling high-end technology from the group into the mass market.

Beyond technical showcases, Geely also brought Zeekr 9X, Galaxy M9, and Gold Brick battery tech achievements. Geely revealed G-ASD will be updated this year with highway L3 and low-speed L4 functionality where regulations allow, including Robotaxi operations.

Great Wall Motors chose a dual breakout: “mechanical + intelligent.” Great Wall launched the Hi4-Z longitudinal twin-motor hybrid architecture, leveraging its full industry-chain R&D prowess, directly addressing users' perennial concerns over range and energy consumption.

Additionally, Great Wall displayed the Soul S2000 horizontally-opposed eight-cylinder motorcycle, the world’s first with a horizontally-opposed eight-cylinder 2000cc engine—a pinnacle of global motorcycle technology. It’s not just mechanical artistry but a silent declaration: Even in internal combustion, a century-long Western domain, Chinese automakers can achieve top craftsmanship.

On another front, Great Wall introduced the ASL 2.0 (Agent of Space & Language) intelligent agent, based on Great Wall’s in-house Coffee EEA 4.0 architecture. This “central computing + regional control” super-neural system gives clear division of labor to the “big brain” and “small brain,” breaking the information barrier between cockpit, intelligent driving, and power.

Driven by AI OS, ASL 2.0 spawns two core intelligent agents: the “Smart Space Agent”—no longer just executing commands, but offering proactive services as a “new family member” via SOA architecture; and the “Intelligent Driving Agent,” which combines VLA and the world model for human-like causal reasoning. This “old driver” cognitive logic allows machines to genuinely understand the physical world with both IQ and EQ.

From concept debut at CES 2025 to 2.0 realization at CES 2026, Great Wall’s rapid pace demonstrates its commitment to cutting-edge tech investment.

Outside OEMs, Chinese supply chain and emerging companies showcased real mass-production capabilities at CES, directly benchmarking against global giants.

Leapmotor showed off its aggressive strategy in electronic-electrical architecture. The newly released Leapmotor D19 uses dual Qualcomm 8797 “Supreme Edition” chips, aiming to challenge the limits of central computing performance. Unlike Sony Honda Afeela’s focus on bringing the PS5 experience to cars, Leapmotor is channeling computing power into vehicle control and real-time inference, emphasizing the hard strength of its “computing base.”

Black Sesame Technologies showcased its A2000 autonomous driving chip, delivering 250 TOPS of single-chip computing power, targeting the L2+/L3 intelligent driving market.

Hesai Technology and Robosense, two LiDAR makers, displayed new-generation products on site. Beyond improved detection distance and precision, their core competitiveness lies in “thousand-yuan-level” cost control.

Continuous Iteration

Through CES 2026, the future trajectory of the automotive industry over the next three years is clearly outlined. Rising technical barriers and changing business models are accelerating industry shakeout.

In the past five years, the threshold for automaking was capital and qualification; in the next five years, it will be “modulus content”—the proportion of AI models in vehicle control.

Advanced intelligent driving is now a costly “data game.” It's not just about algorithms, but also data scale and computing power cost.

Gan Jiayue believes assisted driving now isn’t just about rules or algorithms; maintaining it in the future will require massive resources. Boosting AI capability also demands huge investments of capital, computing power, models, data, and high-quality talent—including accident investigation ability, which requires a large team. It may not be economical for a single company to accomplish “all-stack in-house R&D.”

Geely is sharing high AI R&D costs via scale effects by integrating its internal R&D and building a unified G-ASD base. Gan Jiayue’s “reduce quantity, increase quality” logic reflects this. Only enterprises with annual sales in the millions and massive data feedback can support the training costs of WAM-level large models.

This confirms the industry's forecast: the future intelligent driving market will be dominated by a handful of giants.

For small-to-medium automakers unable to develop in-house large models or unable to afford high-computing-power chips, CES 2026 sends a warning. As “software-defined cars” transition to “AI-defined cars,” technological gaps will quickly lead to market share loss.

Between 2026 and 2027, the industry may see a new wave of mergers and acquisitions. Those unable to achieve “cross-domain integration” like Geely will be forced out.

Moreover, the commercialization of Great Wall ASL 2.0 and Geely Eva’s intelligent agent signifies a fundamental shift in cars’ commercial attributes.

Gan Jiayue points out, future cars are essentially robots. Whether it’s robots or assisted driving, the core is foundational capability: computing power, model capability, and technical data—all three are essential.

The automobile is being redefined as an “embodied intelligent terminal.” When vehicles possess the ability to perceive their environment, understand intent, and even make independent decisions, their value anchor will shift from “horsepower, leather, space” to “computing power, service, experience.”

This means automakers’ valuation logic will change: hardware profit margins will gradually decline, while AI-driven lifecycle service revenues (such as smart driving subscriptions and space services) will become new growth drivers.

This model is disruptive for traditional supply chains. OEMs will reclaim more software control rights, and suppliers must shift from selling parts to selling capability or ecosystem.

Two years ago, Chinese automakers went abroad relying on cost advantages. But at this year’s CES, whether Geely dared to introduce a WAM model “stronger than VLA,” or Great Wall’s definition of “space intelligence,” all demonstrate Chinese automakers are attempting to export technical standards.

The technical confidence Chinese automakers display at CES points to the next stage of global automotive discourse power competition.

Chinese automakers are no longer mere followers of European and American giants, but challengers willing to propose technical routes and define “space intelligence” standards.

In this game, Chinese companies no longer merely need to prove “I can build good cars,” but now confidently tell the world “what a good car of the future should be.”

Risk Warning and DisclaimerThe market has risks; investment must be cautious. This article does not constitute personal investment advice, nor does it consider individual users' specific investment goals, financial status, or needs. Users should consider whether any opinions, views, or conclusions in this article fit their specific conditions. Investments based on this are at your own risk. ```