The 2026 Beijing International Automotive Exhibition opened on April 24 with a new focal point: the rapid ascent of large language models (LLMs) into the smart cockpit.
According to Chinese media reports, Alibaba's Qwen model has formed partnerships with more than 10 automakers, including BYD, Geely, Li Auto, and Changan Automobile. The push underscores how quickly AI systems are being embedded at the center of the in-car experience.
The trend is not limited to domestic brands. BMW has also adopted Qwen as a foundation, integrating it with Banma's in-car AI to introduce a suite of digital agents, including a "vehicle expert," a "travel companion," and a general knowledge assistant. These features are set to debut in the iX3 long-wheelbase model, with the 7 Series to follow.
Meanwhile, ByteDance's Doubao model has already been installed in more than seven million vehicles. At this year's show, it expanded into new partnerships with brands such as Mercedes-Benz, Audi, and Chery's premium line.
Alibaba says its in-car AI assistant will be integrated into cockpit systems across partner brands, enabling drivers to complete multi-step tasks — from route planning and food delivery to hotel booking, ticket purchases, and package tracking — through voice commands. Built on an automotive-grade computing platform from Nvidia, the system is designed to function even in environments with limited connectivity.
At the same time, ByteDance's cloud division, Volcano Engine, unveiled a new generation of automotive AI solutions, offering both a modular cockpit package and a Doubao-based assistant platform. Executives say the system incorporates conversational reasoning, goal-oriented task execution, and adaptive learning — allowing vehicles to move beyond rigid, one-question-one-answer voice interactions.
One notable feature is cross-device integration: the in-car assistant is designed to sync with the Doubao mobile app, with mass production expected later in 2026.
Yet even as AI dominates the narrative, some analysts are questioning whether the idea of the "AI-defined car" risks becoming more slogan than substance.
As automakers rush to embed similar AI capabilities, competition is increasingly framed around whose system performs best. But that convergence is producing a paradox: a proliferation of advanced technologies without clear differentiation. Vehicles are beginning to feel more alike, and for premium brands in particular, the value proposition remains unsettled.
Article translated by Elaine Chen and edited by Jack Wu
