Nvidia has unveiled its new open reasoning vision language model, Alpamayo-R1, at the NeurIPS AI conference, enhancing autonomous driving research. This innovative model is the first of its kind designed for vehicles, enabling them to effectively perceive and analyze their environment. To support developers, Nvidia also released the Cosmos Cookbook, which includes resources for training models. CEO Jensen Huang emphasizes that this technology is pivotal for achieving Level 4 autonomous driving, paving the way for smarter vehicles.