Baidu explores autonomous car rental

Baidu is trying to turn self-driving cars into something closer to a rental car than a robotaxi. That sounds like a small tweak, but it changes the whole business. A robotaxi gives you a ride from A to B. A rental car gives you the car itself for hours or days. On May 8 in Beijing, Baidu Apollo and CAR Inc. — the company behind Shenzhou Zuche — signed a strategic deal to explore exactly that. ### What did they actually announce? The two companies said they will jointly develop an autonomous car rental service. Baidu brings the Apollo self-driving stack. CAR Inc. brings the rental network, fleet operations, and customer base. The framing matters — this was not pitched as another robotaxi expansion. It was pitched as a new category of rental product. ### Why is rental different from robotaxi? (apollo.auto) A robotaxi is basically chauffeuring without the chauffeur. The car picks you up, drops you off, and the operator keeps control of the asset the whole time. Rental flips that. The user keeps the vehicle for a longer stretch and decides where to go inside the service area. That makes the product feel more like “free exploration” than “on-demand transport,” which is why both companies leaned hard on tourism and city landmarks in their pitch. ### Why start with tourism? Because tourism is the cleanest use case. People want flexibility, but they may not want to drive in an unfamiliar city. An autonomous rental car promises the freedom of a short-term rental without the stress of parking, navigation, or local traffic rules. Baidu and CAR said the service could let users move around cultural sites and scenic areas more freely, which is a much easier story to sell than “replace every private car trip.” (apollo.auto) ### What’s the concrete rollout plan? The companies said product definition would begin in the second quarter of 2025, with the service model refined through user feedback and expanded over the next 3 to 5 years. So this was not a same-day mass launch. It was the setup phase — sign the partnership, define the product, test in specific scenarios, then widen the footprint if the economics and safety hold up. (apollo.auto) ### Why does CAR Inc. matter here? Because self-driving software is only half the job. CAR operates one of China’s biggest rental networks, and that means locations, vehicle turnover, cleaning, maintenance, customer verification, and all the annoying logistics that make a fleet usable at scale. Baidu has spent years commercializing Apollo through robotaxis, but rental needs a different operating muscle. CAR already has it. (apollo.auto) ### Is this just a concept, or did it go further? It went further. In July 2025, the two companies said they had officially launched a public autonomous rental product, with reservations opened through the CAR app. Reports at the time described a fully self-service short-term rental flow and said adults 18 and older could book it. That makes the May agreement look less like vague MOU theater and more like the first step in a pretty fast commercialization push. (apollo.auto) ### What’s the hard part now? The hard part is not getting one self-driving car to work. It is making the rental format reliable, safe, and cheap enough to repeat across cities. A robotaxi company controls the trip tightly. A rental model hands users much more flexibility, which means more edge cases, more support needs, and more operational risk. Basically, Baidu is testing whether autonomy can graduate from “book me a ride” to “give me the car.” (tech.cnr.cn) ### Bottom line This deal matters because it widens the commercial target for self-driving cars. Baidu is no longer just chasing the taxi business. It is probing whether autonomous vehicles can become a general mobility product — especially in travel and leisure, where people value freedom more than pure point-to-point efficiency. (apollo.auto)