Nvidia Debuts Agile ONE SN Humanoid, OpenShell at GTC Berlin
TL;DR
- Jensen Huang declared the ChatGPT moment for robotics imminent from Berlin, a deliberate narrative framing targeted at European industrial decision-makers.
- Agile ONE SN delivers 71 total degrees of freedom and 140 Nm torque, with skills transferable to Agile Robots' larger industrial platform.
- Isaac Lab 3.0 ships generally available as open-source, bundled with a 2,000-asset SimReady library and the Human 1M dataset on Hugging Face.
At GTC Berlin, Nvidia introduced Agile ONE SN, a reference humanoid built on Isaac GR00T with a 31-degree-of-freedom body, pitched as a drop-in starting point to cut researcher setup time from weeks to days. The company said it will be available next year.
"This is a great starting point," Jensen Huang told the Tempodrom audience, framing the reference design as modular so researchers can take the full workflow or pull in selected components. The platform leans on Nvidia's existing physical AI stack: Isaac Capture for data collection, GR00T models for robot intelligence, Isaac Lab for simulation-based training, and Isaac OS for runtime. Partners already building on the same platform include Agility Robotics, Figure, KUKA and NEURA Robotics, a lineup that fills much of our recent robotics coverage.
The second thread was agent security. Nvidia announced OpenShell, an open-source, policy-controlled runtime with kernel-level sandboxing designed to isolate what an AI agent is allowed to touch. "The most important thing is to make sure the agent is secure," Huang said. "Every single time an agent is launched, the privileges are different."
Alongside OpenShell came the Agent Knowledge Foundation, a set of models and libraries for connecting agents to enterprise data, including new Nvidia Kumo models for making predictions from structured data. "Ultimately, what turns these super agents, these agents into super agents, is that you give them access to your company's knowledge," Huang said.
The keynote closed with a pitch to European buyers: start from an open model like Nemotron, post-train on proprietary data, and run what Huang called "sovereign intelligence."
What others are reporting
-
Nvidia Investor Relations Read →
Nvidia's canonical press release; the authoritative source record for all product specs, partner OEM list, and Jensen Huang's ChatGPT-moment quote.
"General-purpose robotics is advancing at extraordinary speed — the ChatGPT moment for robotics is fast approaching." — Jensen Huang, Nvidia CEO
-
Agile Robots Read →
First-party from the hardware maker; adds CEO quote, full 71 DOF spec, 140 Nm torque detail, and the skills-transfer path from SN research body to full-scale industrial Agile ONE.
"Agile ONE SN gives R&D labs around the world a complete humanoid platform that is ready from day one." — Dr. Zhaopeng Chen, CEO, Agile Robots
-
Counterpoint Research Read →
Analyst Neil Shah frames GTC Berlin as Europe's shift from pilots to production, then flags the structural risk: every layer of the open platform still runs on Nvidia silicon.
"The stack underneath, from CUDA and NLink to BlueField and Thor, still comes from one vendor." — Neil Shah, Counterpoint Research
-
Unite.AI Read →
Editorial coverage adding real-world deployment context (Bellboy's Billie hotel robot across Marriott and Hilton), TUV SUD certification pathway, and Schaeffler actuator supply-chain detail.
-
The AI Insider Read →
German-focused trade outlet highlights Agile Robots' Munich provenance and an integrated workflow that consolidates six pipeline stages into one system, cited as a key competitive differentiator.
-
Storage Newsletter Read →
Trade wire reprint that explicitly names the Human 1M dataset publication on Hugging Face and three open-source DGX Station GR00T reference workflows.
Originally reported by blogs.nvidia.com
Read the original article →Original headline: Nvidia Expands Isaac With Agile ONE SN Humanoid Reference, Isaac Lab 3.0, and OpenShell Agent Sandbox