Recently, my colleagues and I test drove XPENG's second generation VLA model alongside the latest versions from global peers across China, North America, Europe, Australia, and other regions. My direct impression is that on standard roads, such as main thoroughfares, XPENG's second generation VLA model is already on par with top-tier global products. In complex scenarios like navigating narrow streets, industrial parks, and parking lots, the experience is even better, supporting seamless navigation directly to a parking spot, both efficiently and safely. We are accelerating the global deployment of the second generation VLA model. Recently, my team and I completed localized validation in Germany. With virtually no additional localized training data, the actual driving experience in Germany was remarkably close to that in China. I believe this demonstrates a clear lead over our peers in multi-region global generalization. We aim to be the first to secure regulatory approval in Europe in 2027 and begin delivering the second generation VLA model to more international customers overseas. We look forward to competing head-to-head on the same stage with the most advanced smart driving systems from global peers. XPENG's second generation VLA model was trained in "devilishly complex road conditions." We began by tackling the most complex data and scenarios, aiming not only for superior performance across various operating conditions but also for a single model to better adapt to the road environments and capability requirements of different countries and regions. Solve the hard problems first; start with the end in mind. Ultimately, we must face real global roads and real users, so right from the start, we chose to address genuine, hard-core challenges in complex environments. Last year, Xianming and I made an interesting bet. Today, I want to say, Congratulations, everyone! The Silicon Valley canteen is coming! 🎉 It hasn't been an easy journey. Thank you all for your persistent dedication and hard work.