Nvidia launches Alpamayo, open AI models that allow autonomous vehicles to ‘think like a human’
Summary
At CES 2026, Nvidia launched Alpamayo, a new suite of open-source AI models, simulation tools, and datasets aimed at training physical robots and vehicles to reason through complex driving situations. CEO Jensen Huang stated this marks the "ChatGPT moment for physical AI." The core component is Alpamayo 1, a 10-billion-parameter vision language action (VLA) model that uses chain-of-thought reasoning to help autonomous vehicles (AVs) solve rare edge cases, such as navigating a traffic light outage, by breaking down problems and selecting the safest path. The underlying code is available on Hugging Face for developers to fine-tune or build upon. Nvidia is also releasing an open dataset with 1,700 hours of diverse driving data, and AlpaSim, an open-source simulation framework available on GitHub for safely testing AV systems at scale using synthetic data generated by Nvidia's Cosmos world models.
(Source:TechCrunch)