1. Nvidia introduced Alpamayo-R1, an open reasoning vision language model for autonomous driving research, at the NeurIPS AI conference on December 2nd. This is the first vision language action model specifically designed for autonomous driving, enabling vehicles to perceive and make decisions based on their surroundings.

2. Alpamayo-R1 is built upon Nvidia's Cosmos-Reason model, a decision-making model that processes information before responding. The company released the initial version of the Cosmos model family in January 2025, with additional models following in August.

3. Alongside Alpamayo-R1, Nvidia also provided developers with the Cosmos Cookbook on GitHub. This includes guides, resources, and workflows to help developers better use and train Cosmos models for specific use cases, such as data curation, synthetic data generation, and model evaluation. These releases are part of Nvidia's push into physical AI, which they view as the next frontier for their advanced AI GPUs.