Home: Motoring > Li Auto and XPeng Advance ADAS Upgrades to Rival Tesla's FSD V14

Li Auto and XPeng Advance ADAS Upgrades to Rival Tesla's FSD V14

From:Internet Info Agency 2026-06-17 19:08:00

On June 15, Li Auto announced that its intelligent driving system, "Mach VLA," will launch a new version in the third quarter and aims to achieve capabilities on par with Tesla’s FSD V14 by the fourth quarter. The system leverages Li Auto’s self-developed large models—Mach Mind-Pro and Mach Mind-Edge—for language understanding and logical reasoning, while Mach VLA handles 3D visual perception and vehicle motion control. Li Auto has also begun mass production of its in-house developed chip, the Mach M100, completing full-stack self-reliance across chips, compilers, operating systems, AI algorithms, and domain controllers. XPeng stated that its second-generation VLA physics-based AI large model is expected to match the overall performance of Tesla’s FSD V14.2 in Silicon Valley by August. This model directly maps visual inputs into vehicle action commands, eliminating intermediate processing steps and reducing intelligent driving response latency to under 80 milliseconds. The system is optimized for XPeng’s self-developed Turing AI chip, delivering 2,250 TOPS of effective computing power and supporting local onboard deployment of models with billions of parameters. Tesla’s FSD V14 features a restructured neural network architecture based on an autoregressive Transformer model, which reduces road condition recognition latency and enhances decision-making accuracy in complex scenarios. FSD has already received public road approval in five EU countries—including the Netherlands, Denmark, and Belgium—and has been rolled out in South Korea and Australia. In China, Tesla has rebranded the system as “Tesla Assisted Driving,” with its supervised version now available in 10 countries and regions, including China. Tesla has also established a local AI training center in China, deploying localized training capabilities to prepare for broader future deployment.

Editor:NewsAssistant