Autohome launches AI agent 'Zhishi Che Jiaguan' spanning car selection, purchase, ownership and resale

Autohome held its AI product launch in Beijing on September 5 under the theme "Car matters? Ask Zhishi" (车的事，找芝士), unveiling Zhishi Che Jiaguan (literally "Cheese Car Butler"), an AI agent designed to connect the entire vehicle lifecycle — selection, purchase, ownership and resale — in one natural-language experience. The event gathered AI ecosystem partners, new-retail and used-car businesses, car-service merchants and owner representatives, which Autohome framed as its strategic move into vertical AI services for the automotive industry.

Chairman and CEO Liu Chi said AI is not just an upgrade in technical capability but a redefinition of how car services are delivered, adding that more than two decades of accumulated automotive content, data and service experience are being converted into the ability to solve real user problems. CTO Xiang Bibo distilled the product logic into two layers: trustworthy decisions for car selection, and reliable services for everyday car use.

On the selection side, the concept is "AI capability + real car-owner ecosystem": AI understands needs, consolidates information and speeds up decisions, while real owners contribute first-hand experience that public information cannot cover. Product lead Chen Xi stressed that an AI agent should be judged not by how many questions it can answer, but by whether it delivers a clear recommendation and keeps fulfilling service when follow-up needs appear — for example, when a user hesitates between two models, Zhishi weighs space, range, charging and family use together and offers a directional recommendation with the reasoning behind it.

On the ownership side, needs such as maintenance, repairs, charging, insurance and roadside rescue are infrequent and fragmented, so the agent pairs AI with a service ecosystem that can actually fulfill requests, compressing information judgment, service matching and real delivery into one path. Looking ahead to the gradual arrival of L4-level autonomous driving, Autohome expects cars to shift from "waiting for a human to initiate service" to "autonomously identifying and invoking service," with capabilities extending from serving owners to serving the vehicle itself and reserving interfaces for autonomous charging, washing, repair and maintenance.

The experience rests on a dual foundation of an industry model and an industry ecosystem: Autohome's self-developed Cangjie automotive large language model handles intent understanding, vehicle knowledge and decision support, while brands, dealers and finance, insurance, maintenance and repair resources deliver services offline. Zhishi Che Jiaguan is already live on the Qwen, Xiaoyi and WorkBuddy platforms to broaden its entry points across different intelligent ecosystems.

On the ecosystem side, Autohome is drawing on its base of 150 million accumulated car owners, turning their first-hand experience into a knowledge source for the AI and encouraging contributions through benefit, interaction and honor incentives, with top contributors potentially earning perks such as free vehicle use for a year. On the merchant side, it has connected nearly 30,000 partner 4S dealerships, more than 12,000 used-car dealers and over 100,000 car-service resources across 369 cities, backed by admission mechanisms including qualification checks, capability verification and credit ratings.

That is one of the sharpest differences between vertical automotive AI and general-purpose AI: value lies not only in the answer itself, but in whether an executable service follows it. For Autohome, the significance of Zhishi Che Jiaguan lies in reorganizing the relationship between information, judgment and service so AI genuinely enters car-owner life; the open question now is whether the selection-to-resale loop can be closed through the interplay of the owner and merchant ecosystems, and extended to more scenarios and vehicle-machine services in the L4 era.