NVIDIA Releases Alpamayo 2 Super: A 34B Open Vision-Language-Action Model for Robotaxis and Autonomous Driving Under OpenMDW-1.1

MarkTechPost · 2 months ago · 44 ● 2 sources

NVIDIA released Alpamayo 2 Super, a 34-billion-parameter vision-language-action model designed for autonomous driving that combines a language-reasoning backbone with a diffusion-based action decoder. The model achieved a LingoQA score of 79.2, outperforming nearly 40 competitors including GPT-4o by 23.2 points, and was trained on 115,000 hours of multi-camera driving video with 3.7 million causally-linked decision explanations. The weights are distributed under the permissive OpenMDW-1.1 license, allowing commercial deployment and redistribution without additional permission.