NVIDIA has introduced Groot N1, an open-source, pre-trained foundation model designed to enhance the development and capabilities of humanoid robots. Announced during the GTC 2025 keynote by CEO Jensen Huang, Groot N1 is a ‘generalist’ model trained on a combination of real-world humanoid movement data, synthetic data generated with NVIDIA’s Isaac GR00T Blueprint, and large-scale video data from the internet.
The model features a dual-system architecture inspired by human cognitive processes. This architecture includes ‘System 1 (Diffusion Transformer)’ and ‘System 2 (Vision-Language Model).’ Based on NVIDIA-Eagle with SmolLM-1.7B, System 1 acts like human reflexes, enabling fast, intuitive actions. On the other hand, System 2 functions as a vision-language model, allowing the robot to reason about its environment and plan complex tasks.
This combination allows humanoid robots to perform complex tasks (like grasping objects and executing multi-step operations) autonomously, with System 2 planning actions and System 1 executing them through precise movements. “The age of generalist robotics is here,” NVIDIA CEO Jensen Huang said during the event.
Speaking of details, GR00T N1 is a single AI model with one set of weights, meaning it does not rely on multiple models for different tasks. This unified model enables humanoid robots to perform various manipulation behaviours. It is designed to work on robots like the Fourier GR-1 and 1X Neo, which are advanced humanoid robots capable of interacting with objects.
Notably, NVIDIA’s Groot N1 is an evolution of the company’s earlier Project Groot. Early access to Groot N1 has been granted to companies like Boston Dynamics, Agility Robotics, Mentee Robotics, and Neura Robotics.
Meanwhile, NVIDIA has made Groot N1’s training data and evaluation scenarios available on platforms like Hugging Face and GitHub, enabling developers to customize the model further using post-training data. Actually, the AI model is adaptable and can be fine-tuned for different types of robots, specific tasks, and various environments.
The move becomes more significant considering NVIDIA’s major focus on the AI-driven robotics segment. Earlier, reports suggest that the company is launching a compact ‘Jetson Thor’ computer for AI-powered humanoid robots in 2025. Interestingly, as physical AI and robotics are seen as the next big tech revolution, NVIDIA’s push into robotics comes at a time when the company is facing increased competition for its powerful AI chips from rival chipmakers like AMD.
Clearly, after artificial intelligence (AI), humanoid robotics is the new buzz among tech giants. ChatGPT maker OpenAI is planning to develop a full-fledged humanoid robot. Meanwhile, Korean tech giant Samsung has also joined the humanoid robotics race with its acquisition of Rainbow Robotics. Also, companies like Tesla (Optimus) and Boston Dynamics are already working on humanoid robots for household and industrial tasks.
In fact, the humanoid robot market is expected to grow rapidly over the next decade. In 2024, the market is valued at $2.98 billion, which is relatively small. However, by 2035, it is projected to reach a massive $243.40 billion, with a CAGR of 49.21%.
Content originally published on The Tech Media – Global technology news, latest gadget news and breaking tech news.
