NVIDIA's new take on OpenClaw is meant for enterprises
What's the story
NVIDIA has unveiled a new enterprise-grade AI agent platform called NemoClaw. The announcement was made by the company's CEO Jensen Huang during his keynote at the GPU Technology Conference (GTC). The platform is built on OpenClaw, an open-source framework for creating and running AI agents locally on a company's hardware.
Platform features
NemoClaw offers enhanced security and privacy
NemoClaw is essentially an upgraded version of OpenClaw, with added security and privacy features. The platform aims to provide a secure environment where enterprises can control the behavior of AI agents and how they handle data. Huang emphasized the importance of having an "OpenClaw strategy" for every company, similar to strategies for Linux, HTTP, HTML, and Kubernetes in the past.
Partnership
It integrates with NeMo software suite
NVIDIA collaborated with Peter Steinberger, the creator of OpenClaw, to develop NemoClaw. The platform supports any coding agent or open-source AI model, including NVIDIA's NemoTron open models. It also provides access to cloud-based models on local devices and is hardware agnostic, meaning it doesn't have to run on NVIDIA's own GPUs. The platform integrates with NeMo, NVIDIA's suite of AI agent software.
Development phase
NemoClaw is an alpha release
NVIDIA has described NemoClaw as an early-stage alpha release, with some rough edges still to be worked on. The company is working toward a production-ready sandbox orchestration, but for now, the focus is on getting users' own environments up and running. This comes as part of a wider trend in the AI space where building enterprise AI agent platforms has become a major focus in recent months.