The automotive industry's battle for autonomous driving supremacy has already set the stage for next year's competition. On December 11, XPeng Inc. (NYSE: XPEV) Chairman He Xiaopeng issued a bold challenge on social media: If XPeng's VLA technology matches Tesla Motors' (NASDAQ: TSLA) FSD V14.2 performance in Silicon Valley by August 30 next year, he will build a Chinese restaurant in Silicon Valley. If not, XPeng's autonomous driving head Liu Xianming will "run naked" across the Golden Gate Bridge.
He's wager stems from his extensive testing and observation of Tesla's FSD system. After test-driving the latest FSD V14.2 and Robotaxi in Silicon Valley, He noted that Tesla's system has evolved from "smooth L2+ advanced driver assistance" to "near-L4 reliability" within a year. During five hours of testing, he found consistent performance between standard vehicles and Robotaxis using the same model, acknowledging "significant improvements over last year's capabilities."
While praising Tesla's progress, He openly admitted XPeng's VLA currently trails FSD V14.2 in functionality. However, he revealed VLA 2.0 will launch next quarter. Earlier in August, He predicted XPeng's Ultra models would surpass competitors' ADAS capabilities "by dozens of times or more" when its Robotaxi begins trials in 2026, leveraging L4-grade technology.
At XPeng's 2025 Tech Day, the company unveiled its second-generation VLA model, eliminating traditional language translation steps to enable direct vision-to-action processing. This innovation allows navigation-independent autonomous driving globally, with full deployment to Ultra models planned for Q1 2026. He emphasized XPeng's heavy investment aims to enable safe operation across complex environments - from highways to Beijing's hutongs - seeing this as critical for true autonomous driving.
XPeng is upgrading hardware too, equipping Ultra versions of G7 and P7 models launching in Q3 2025 with three self-developed Turing AI chips delivering 2,250 TOPS. The company also plans three Robotaxi models for 2026, partnering with Amap while focusing solely on development, not operations. He projects automakers will need $50 billion annual R&D spending, with $30 billion dedicated to AI.
This challenge reflects Chinese automakers' ambition to rival global leaders in core autonomous tech. With VLA 2.0's launch and Robotaxi progress, XPeng accelerates its pursuit through full-stack R&D. Yet Tesla's rapid iterations remind that this competition extends beyond algorithms to encompass data, computing power, and engineering capabilities. The wager's outcome - whether bringing Chinese cuisine to Silicon Valley or an executive dash across the Golden Gate - will demonstrate if XPeng can truly become China's answer to Tesla.
Comments