NemoClaw

Overview

NVIDIA NemoClaw is an open source stack that adds privacy and security controls to OpenClaw. It enables the deployment of more secure, always-on AI assistants with a single command. NemoClaw uses NVIDIA Agent Toolkit software to secure OpenClaw, installing components to enforce policy-based privacy and security guardrails, giving users control over how agents behave and handle data. It evaluates available compute resources to run high-performance open models locally for enhanced privacy and cost efficiency.

Features

• Run Claws More Safely: Adds security and privacy controls to OpenClaw, allowing developers to build and run AI assistants with more confidence.
• Use Any Coding Agent: Supports open models locally on dedicated systems, with a privacy router connecting agents to cloud-based frontier models within defined privacy and security guardrails.
• Deploy Anywhere: Provides local, 24/7 compute for autonomous agents on dedicated systems to run tools, write code, and complete tasks.

Use Cases

• Deploying always-on, self-evolving autonomous agents safely and instantly.
• Running high-performance open models locally for enhanced privacy and cost efficiency.
• Building trustworthy AI agents with enhanced safety and privacy using OpenShell and AI-Q for reasoning agents that transform enterprise data into explainable results.

Getting Started

Install with the command: $ curl -fsSL https://nvidia.com/nemoclaw.sh | bash Follow up with: $ nemoclaw onboard