In early January 2026, a hint of chill still lingers in the morning breeze of Las Vegas, yet the Las Vegas Convention Center is already abuzz with excitement.

Venture into the automotive technology hub at CES during this period, and you'll notice a subtle yet significant transformation: while the booths of traditional international automakers maintain their elegance, restraint, and historical gravitas, it's the booths adorned with Chinese brands that truly captivate and repeatedly draw the crowds' attention.

Reflecting on the past two or three years, discussions about intelligent automotive technologies were often centered around questions like, 'Is the screen large enough?', 'Is voice recognition responsive enough?', and 'Does highway NOA (Navigate on Autopilot) run smoothly?'. Back then, intelligence seemed like a checklist of features—the more items and higher parameters, the easier it was to become a selling point.

However, by 2026, this narrative has clearly evolved. Intelligent automotive technologies are now delving into a more challenging and authentic system-level exploration.

So, at this critical juncture where showmanship gives way to true implementation, how far has Chinese automotive technology truly advanced? The answer may be further than many imagine.

From Showcasing Features to Competing on Systems

Several years ago, competition in intelligent automotive technologies was still at a superficial stage, focusing on 'visibility and tangibility'. At major auto shows and product launches, manufacturers were eager to showcase intuitive and quantifiable technological highlights: Can voice assistants engage in multi-round natural conversations? Is the central control screen larger and more dazzling than competitors'? Has urban NOA coverage extended to thousands of kilometers?

As long as features were quickly piled on and parameters pushed to new heights, media attention and consumer interest could be captured in a short time. Back then, intelligence was more of a marketing label, a quick card for product differentiation. Automakers, akin to assembling a jigsaw puzzle, pieced together the latest and most dazzling intelligent configurations, pursuing short-term exposure and sales conversion.

However, by the time of CES 2026, this feature-centric discourse system has clearly taken a backseat. Booths are no longer filled with repeated emphasis on single technological metrics; instead, keywords like 'system architecture', 'full-domain collaboration', 'intelligent agents', and 'central computing' with greater engineering depth and strategic height have taken center stage.

Among this shift, the actions of leading Chinese automakers are particularly noteworthy.

Geely introduced a technology system called Full-Domain AI 2.0 at this year's exhibition, centered around a World Behavior Model (WAM). The name may sound abstract, but it points to a very specific goal: creating a unified decision-making hub for the entire vehicle. In the past, intelligent driving, cabin interaction, and chassis control operated independently on different computing units, with disconnected data and fragmented logic.

WAM attempts to break down these barriers, using a multimodal large model to understand users' vague intentions and then, through a lower-level action derivation module, coordinating intelligent driving route planning, energy system range assessment, and ultimately forming a coherent set of services. Around this model, Geely has built a collaborative framework centered around the Vehicle Intelligent Agent (VIA) Eva, linking intelligent driving and cabin as two main pillars and extending to multiple subsystems such as chassis and thermal management, truly achieving a transition from single-point feature breakthroughs to system-driven innovation.

Great Wall Motors, on the other hand, chose to reconstruct the human-vehicle relationship from a service logic perspective. Its ASL Intelligent Agent System aims to transform the vehicle from a passive tool executing instructions into a partner capable of actively understanding needs. Traditional interactions require users to specify 'increase the air conditioning temperature', while ASL hopes to handle vague yet authentic expressions like 'I'm a bit cold'. To support this capability, Great Wall Motors deployed an upgraded VLA (Vision-Language-Action) large model on the vehicle side and clearly defined the responsibilities between the cloud and the vehicle: the cloud handles extreme scenario generation for training through simulation, while the vehicle focuses on safe and efficient real-time decision-making. It is reported that mass-produced models equipped with this system will hit the market in June 2026, once again demonstrating the rapid technological mass production capabilities of Chinese automakers.

On another front, Leapmotor collaborated with Qualcomm to further promote the integration of electronic and electrical systems at the hardware architecture level. Its central computing platform, developed in collaboration with Qualcomm, consolidates functions such as intelligent cockpit, intelligent driving, body control, and vehicle gateways, which were originally dispersed across different controllers, into a single hardware unit driven by dual Snapdragon 8797 chips. On this platform, the cockpit's interaction model and the intelligent driving's perception model can run in parallel without interference. Meanwhile, through a service-oriented software architecture, over 200 combinable capabilities are opened up, providing a foundation for personalized experiences. This platform is confirmed to debut on Leapmotor's upcoming D19 model, becoming one of the earliest mass-produced vehicles globally to adopt such a cross-domain integration solution.

It can be said that intelligent automotive technologies in China have transitioned from showcasing features to competing on systems, with intelligence being redesigned as an integral part of the vehicle's capabilities. This is precisely the natural choice after technology returns to rationality and enters the deep waters.

From Concept to Application

If 'system' is one keyword at CES 2026, then 'implementation' is another. Unlike previous years, when there was enthusiasm for showcasing cutting-edge technology prototypes and emphasizing 'disruptive innovation', practical terms like 'mass production', 'compliance', 'overseas availability', and 'cost controllability' were repeatedly mentioned at this year's exhibition. Multiple participating companies, in addition to showcasing products, simultaneously provided information on model integration ranges, launch timelines, and commercialization paths. This means that more and more technologies have moved from the laboratory or concept demonstration stage to the broader real world.

In the field of autonomous driving, Geely's G-ASD system, iterated based on data from 8.5 million intelligent vehicles, has been deployed across 16 models from brands like Zeekr and Lynk & Co, covering over 300,000 vehicles. It is reported that Geely is simultaneously advancing operational testing related to Robotaxi, planning to cover high-speed autonomous driving Level 3 and low-speed autonomous driving scenarios Level 4 by 2026, preparing for large-scale commercialization.

Around system-level intelligence, software and cabins have become another clear path for implementation. BOE's global debut of HERO 2.0 Intelligent Cockpit at CES achieved a 98% recognition rate for complex instructions; Geely Automobile's EVA Hyper-Humanoid Emotional Intelligent Agent, driven by the Xingrui Large Model, boasts a 99.8% wake-up success rate, a 0.2-second response speed, and supports six-zone interaction, having already been mass-produced in the Galaxy M9.

The implementation of intelligence cannot be separated from the fundamental support of vehicle platforms and energy systems. At the vehicle and powertrain system level, Great Wall Motors chose to let its products speak, showcasing a full-stack implementation path from underlying hardware to complete vehicle products. Its exhibited semi-solid-state batteries, hydrogen fuel cell engines, and the Hi4-Z architecture have constructed a diversified energy matrix covering hybrid, pure electric, and hydrogen energy. The unveiling of models such as the Wey Highland 9, Tank 500 Hi4-T Intelligent Edition, and Soul Motorcycle S2000 CL further demonstrates its implementation capabilities. Among them, the Tank 500 Hi4-T Intelligent Edition integrates a non-decoupled four-wheel-drive structure with full-scenario NOA, balancing the characteristics of off-road capability with the intelligence of high-level autonomous driving.

Similarly, the maturity of the upstream industrial chain provides solid support for vehicle implementation. Chinese LiDAR supplier Hesai Technology debuted the ETX/FTX dual LiDAR combination at CES 2026 and simultaneously announced that it would double its annual production capacity to 4 million units in 2026. Moreover, Hesai's new-generation L3 automotive-grade LiDAR solution has also secured its first mass production fixed point for passenger vehicles, with plans to initiate mass production by the end of 2026 or early 2027.

From autonomous driving and cabin systems to perception hardware, these real industrial plans collectively reflect a trend: intelligent vehicles are beginning to leave the laboratory and enter a mature stage truly oriented towards users and markets. This integrated capability of edge-cloud unity gives Chinese automakers significant advantages in handling complex urban scenarios, addressing long-tail problems, and enabling continuous OTA (Over-the-Air) evolution.

The Discourse Power in Intelligent Driving Is Shifting Eastward

A detail easily overlooked is the subtle change in the influence of intelligent automotive technologies in China at CES 2026.

For a considerable period, the narrative of innovation in intelligent automotive technologies was almost entirely dominated by European and American countries. Whether it was the route of high-level autonomous driving relying on high-definition maps or the cabin experience paradigm led by luxury brands, the industry's default premise was always that technology definitions came from Europe and America, and Chinese automakers were more followers and appliers.

However, on the exhibition floor of CES 2026, this long-standing entrenched cognition is quietly loosening. Intelligent automotive technologies in China are beginning to be incorporated into broader technological discussions as referenceable and replicable solutions.

One of the most frequently mentioned directions is the 'perception-heavy, map-light' autonomous driving technology route. For a long time, European and American premium brands have generally relied on high-definition maps to support high-level autonomous driving, but these maps suffer from long update cycles, limited coverage, and high maintenance costs. Today, the map-free solutions promoted by Chinese automakers such as Huawei and Geely have been truly implemented in China's complex urban road conditions, with high-level autonomous driving systems like ADS 4.0 even being introduced into overseas premium models such as Audi. It can be said that the 'perception-heavy, map-light' or even 'map-free' technology routes explored by Chinese automakers are no longer seen as compromises made under resource constraints; instead, they have become new references for global autonomous driving evolution due to their stronger generalization capabilities and lower operation and maintenance costs.

Similar reversals are also occurring in the field of intelligent cabins. In the past, experience paradigms such as HUD (Head-Up Display) and human-machine interaction were mostly proposed by German or American overseas luxury brands. However, today, some overseas luxury models are beginning to incorporate software architectures, interaction systems, and computing platforms dominated by the Chinese supply chain, designing cabins as vehicle-level intelligent systems. Chinese supply chain companies such as BOE, Desay SV, and Foryou Group are bidding farewell to being mere component suppliers and are beginning to participate in early definitions as designers of vehicle-level intelligent systems.

This upgrade in influence means that Chinese automakers have broken free from the Western-dominated technological framework and are beginning to participate in the competition for intelligent automotive technologies with their own architectures.

For the automotive industry, CES 2026 may not be a cross-generational singular moment or a decisive victory, but it serves as a periodic display, clearly presenting the accumulations of intelligent automotive technologies in China over the past few years on the international stage.

It is foreseeable that with the formal legalization of Level 3 autonomous driving in China and the accelerated advancement of Robotaxi commercialization pilots, the intelligent advantages of Chinese automobiles will be further amplified. CES 2026 may well be the critical node for the reshaping of the global automotive industry's power structure.