Chongqing Afari Technology (601777.SH) has released a new generation of intelligent driving technology, in collaboration with the super intelligent body, focusing on the integration of cabin and driving.

Cockpit integration is becoming the core trend in the development of intelligent cars in 2026. At the NVIDIA GTC 2026 conference held today, GEELY AUTO Group announced a major breakthrough in the field of "cockpit integration" and will soon release and mass produce the "Super Eva + G-ASD 4.0" jointly developed with Chongqing Afari Technology (601777.SH) and Jieyue Xingchen. The Xpeng P8X will be the first in the industry to be equipped with the world's first intelligent cockpit, intelligent driving assistance, and digital ecosystem that integrates all aspects of the super intelligent body, enabling dialogue between the cockpit and driving, and comprehensive digital ecosystem service linkage. The Super Eva is jointly developed by GEELY in cooperation with Chongqing Afari Technology and Jieyue Xingchen. Based on a large model base, it is a vehicle-level intelligent body that integrates intelligent cockpit, intelligent driving assistance, chassis, power, and other underlying systems natively. Its technological concept is highly compatible with the conversational control architecture represented by Tesla's Grok. The Thousand Li Hao Han G-ASD, jointly developed by GEELY and Thousand Li Zhijia, is an advanced driver assistance solution with high "modularity", which was officially released for mass production in January this year and completed rapid iterations from version 3.0 to 4.0 in just two months. Its technical path is similar to Tesla's FSD solution. Under the unified concept of cockpit integration, Super Eva and Thousand Li Hao Han G-ASD achieve technological convergencea super intelligent car partner with more "human-like" characteristics is about to debut. The Super Eva deeply integrates cutting-edge technologies such as Agentic AI and end-to-end models. After installing Super Eva, users will experience a revolutionary human-car interaction innovation, with core functions such as deep chat, deep navigation, external vision, and digital ecosystem fully activated. With just one sentence, Eva can transform into a personal super assistant, able to think independently, deeply understand user intentions, remember context and user preferences, and automatically plan and execute tasks based on vague instructions. For example, just say "Take me to pick up my child from school, find a McDonald's on the way, I need to be at the school by 5 o'clock", and Eva will accurately understand the vague semantics, autonomously plan the navigation route, start intelligent driving assistance, and complete operations such as ordering meals on the way, stopping to pick up meals, and automatically parking at the school gate. Compared to traditional voice assistants that only superficially integrate models with existing in-car systems, Super Eva brings a more human-like intelligent and interactive experience. It is no longer just a "chatting" in-car assistant but a super intelligent car body that understands you better and can get things done. Thousand Li Hao Han G-ASD 4.0 adheres to the high modularity route, making the driving experience safer and smoother with model-driven assistance, like an experienced "AI old driver" behind the wheel. First, the unified model brings a more humanized car control experience. When maneuvering through narrow lanes, it can accurately identify oncoming vehicles and obstacles, adjust speed and trajectory in real-time; in scenarios like turning at T-junctions and entering main roads from secondary roads, the driving trajectory is stable and smooth, with almost no unnecessary steering corrections; precise dynamic predictions for motor vehicles and non-motor vehicles make interactions more natural and confident. Secondly, based on the large model's physical world common sense, G-ASD 4.0 has stronger semantic understanding capabilities of scenarios. In high-speed tollgate scenarios, it can autonomously judge and accurately enter manual or ETC lanes without needing any intervention. Especially facing more complex manual lanes, it can precisely control the distance to the side, smoothly stop at the window position, quickly recognize the signal to lift the barrier after completing the operation, smoothly accelerate and depart; in unfamiliar underground parking lots, the vehicle can actively find the exit and roam without having a map. It is not just about finding spaces to move through but also able to understand directions, cross-level ramps, identify gate entrances and exits, clearly distinguish between ground and underground scenes, offering a truly "purposeful and human-like" intelligent roaming experience. In the future, the key to highly integrated cockpit technology lies in the integration and synergy among intelligent cockpits, intelligent driving, chassis, power, and other systems. As one of the industry's few Tier 1 platform enterprises driven by AI large models and possessing cockpit and driving solutions capabilities, Chongqing Afari Technology is working with partners to jointly define the next generation of intelligent travel product experiences and create a collaborative paradigm with more imagination space for the transformation of AI cars.