Nvidia has launched Alpamayo 2 Super, a 32-billion-parameter reasoning‑based vision language action (VLA) model that extends the Nvidia Alpamayo family of open AI models, simulation frameworks and physical AI datasets for safe Level 4 robotaxi development.
The company has also announced new tools, models and agent skills that complete the pipeline from real-world data capture to closed-loop training and in-vehicle deployment, including Nvidia AlpaGym, Nvidia OmniDreams and new Nvidia Omniverse NuRec models.
According to Nvidia, Alpamayo 2 Super helps accelerate autonomous vehicle (AV) development by eliminating the need to build key autonomy infrastructure from scratch. It reportedly enables humanlike perception, reasoning and action, and provides the interpretability needed for safety validation and regulatory collaboration.
To better train models for on-road deployment, the AlpaGym framework provides a platform for closed-loop reinforcement learning (RL). The Nvidia OmniDreams generative world model for photorealistic closed-loop AV scenario generation enables developers to simulate rare and long-tail driving scenarios at scale.
Nvidia is providing physical AI agent skills for all of its AV development tools. For example, the Neural Reconstruction skill powered by Nvidia Omniverse NuRec uses real-world fleet driving scenarios for simulation, and generates synthetic training data at scale.
“Alpamayo is the moment cars begin to safely reason, not just drive,” said Jensen Huang, the founder and CEO of Nvidia. “Only Nvidia makes available open models, simulation, real-world data and agent skills so the entire global robotaxi ecosystem can develop Level 4 capabilities that understand edge cases, explain decisions, earn trust and scale safely to millions of vehicles.”
Reasoning-based AVs
The Alpamayo family now scales from 10 billion to 32 billion parameters with Alpamayo 2 Super, going beyond trajectory generation to reason, plan and act across the full driving stack, according to Nvidia. With multitask capabilities spanning reasoning, auto-labeling, scene understanding, model critiquing and distilling knowledge into smaller models, it provides the building blocks for scalable L4 AV development and deployment.
Designed as a teacher model, Alpamayo 2 Super can be distilled into compact models that run on the accelerated compute of the Nvidia Drive Hyperion platform – Nvidia Drive AGX Thor, which runs inside the vehicle.
As the teacher model scales from 10-billion-parameter models like Nvidia Alpamayo 1 Nano and Nvidia Alpamayo 1.5 Nano to 32 billion parameters with Alpamayo 2 Super, a downstream AV stack built on Alpamayo reportedly inherits higher‑quality reasoning and perception from a single open release, without requiring each manufacturer to rebuild from scratch.
Closed-loop training and deployment cycles
Nvidia AlpaGym is an open-source, high‑throughput, closed‑loop RL framework. Where open‑loop training evaluates models against recorded data and generates a single round of actions, AlpaGym runs models through continuous decision and observation cycles in Nvidia AlpaSim, with every braking, steering and navigation choice affecting the environment.
As a result, the company says that AlpaGym exposes the compounding errors and edge‑case failures that static datasets miss, allowing models to learn from experience.
Built on the AlpaSim microservice simulation stack and Nvidia Omniverse NuRec, AlpaGym reportedly enables efficient, scalable, closed-loop RL to push the frontier of driving performance. In combination with the Physical AI AV Dataset, Alpamayo provides a continuous path from open-loop pretraining to closed-loop refinement.
Nvidia is also releasing the CoC Auto-Labeling Pipeline as open source on GitHub. The pipeline automatically generates decision-grounded and causally linked CoC labels from raw driving clips with no human annotation required, providing the causal training data foundation needed to train embodied reasoning models at scale.
Physical AI agent skills for AV
To support reasoning-based AV development, Nvidia is launching new physical AI agent skills as part of Nvidia Agent Toolkit, to guide developers and their coding agents through the simulation, data generation and closed-loop training workflows needed to build and validate autonomous driving systems at scale. This includes Neural Reconstruction skills powered by Nvidia Omniverse NuRec libraries, Nvidia OmniDreams skills for photorealistic scenario generation and AlpaGym skills for closed-loop RL.
In related news, next-generation Helm.ai models deliver full-HD 360° synthetic driving environments
