NVIDIA announced that ISAAAC GR00T N1-OPEN SOURCE, a claimant, but a configurable foundation model, which is aimed at accelerating the development and capabilities of humanoid robots-is now available. “The age of generalistic robotics is here,” says the founder and general director of Nvidia, Jensen Huang. “Thanks to NVIDIA Isaac GR00T N1 and new data generation frames and a robot learning, robotics programmers will open the next border in the AI era everywhere.”
During today’s speech GTC 2025 Huang showed Humanoid Robot Neo Gamma 1x Performing autonomous cleaning tasks using after training a policy built on the GR00T N1 model. “The future of humanoids concerns adaptation and learning abilities,” says Bernt Børnich 1x. “The NVIDIA GR00T N1 model is a serious breakthrough for the reasoning and ability to reason robots. With a minimum amount of data after the training, we were able to fully place on the Neo Gamma-by saving our mission to create robots that are not tools, but companions that can help people in significant, immeasurable ways.”
Other companies developing humanoid robots that had early access to the GR00T N1 model are Boston Dynamics, creators of Atlas; Robotics of agility; Robotics mentee; And neura robotics.
System 1, as NVIDIA calls, is described as a “fast -thinking model of action”, which behaves similarly to human reflexes and intuition. It has been trained in the scope of data collected through interpersonal demonstrations and synthetic data generated by the Omniverse Nvidia platform.
System 2, which is powered by a vision language model, is a “slow model”, which “the reasons for its environment and the instructions he received to plan activities.” These plans are transferred to System 1, which translates them into “precise, constant robot movements”, which include gripping, moving objects with one or two arms, as well as more elaborate multi -stage tasks including combinations of basic skills.
While the GR00T N1 foundation model is subject to generalized humanoid reasoning and skills, developers can adapt their behavior and the possibilities of specific needs, after training with data collected from demonstration or interpersonal simulations.