Nvidia said it is teaming with partners to ignite the next era of AI with open source software for autonomous, self-evolving enterprise AI agents — increasing agent safety, security and efficiency — to speed a generational shift in software and knowledge work.
The Nvidia Agent Toolkit provides open source models and software for enterprises and developers building tools that scale productivity by autonomously determining how to complete assigned tasks. It now includes Nvidia OpenShell — an open source runtime that enforces policy-based security, network and privacy guardrails that make autonomous agents, or claws, safer to deploy.
Nvidia announced the news during the GTC keynote by CEO Jensen Huang at the company’s GTC event on Monday in San Jose, California.
“Claude Code and OpenClaw have sparked the agent inflection point — extending AI
beyond generation and reasoning into action,” said Huang, in a statement. “Employees will be supercharged by teams of frontier, specialized and custom-built agents they deploy and manage. The enterprise software industry will evolve into specialized agentic platforms, and the IT industry is on the brink of its next great expansion.”
Nvidia Agent Toolkit speeds community and enterprise software evolution
Nvidia Agent Toolkit includes open models like Nvidia Nemotron, open agents like Nvidia AI-Q, open skills like Nvidia cuOpt and open runtimes like OpenShell.
Developers can use Agent Toolkit software for creating specialized AI agents that can act autonomously, while using and building other software to complete tasks.
Using the NVIDIA AI-Q open agent blueprint, developers can create custom AI agents that can perceive, reason and act on enterprise knowledge, automatically choosing the right data sources and depth of analysis to deliver precise, context-aware answers. A built-in evaluation system explains how each AI answer is produced.
The AI-Q hybrid architecture uses frontier models for orchestration and Nvidia Nemotron open models for research, which can cut query costs by more than 50% while providing world-class accuracy. Nvidia used the AI-Q Blueprint to develop the top-ranking AI agent on the DeepResearch Bench and DeepResearch Bench II leaderboards.
Nvidia OpenShell gives autonomous agents the access they need to be productive while enforcing policy-based security, network and privacy guardrails. Nvidia is collaborating with security providers including Cisco, CrowdStrike, Google, Microsoft Security and TrendAI to build OpenShell compatibility with their cyber-and AI-security tools.
Agent engineering company LangChain — which provides open source frameworks that have been downloaded over 1 billion times — is working with Nvidia to integrate Agent Toolkit, including AI-Q, OpenShell and Nemotron open models, datasets and tools, into the LangChain deep agent library for developing advanced, accurate enterprise AI agents and running them at scale.
Software leaders across industries are working with Nvidia Agent Toolkit software including OpenShell and Nvidia Nemotron open models to expand the capabilities of AI agents in their applications and platforms.
Nvidia announced partnerships with Adobe, Amdocs, Atlassian, Box, Cadence, Cisco, Cohesity, CrowdStrike, Dassault Systèmes, IQVIA, Palantir, Red Hat, Salesforce, SAP, Siemens, ServiceNow and Synopsys.
Developers can explore Nvidia Agent Toolkit and OpenShell on build.nvidia.com
today, running on inference providers and Nvidia Cloud Partners including Baseten,
Bitdeer AI, CoreWeave, DeepInfra, DigitalOcean, GMI Cloud, Fireworks, Lightning,
Together AI and Vultr.
Developers can also use OpenShell with LangChain, and download it from GitHub to
run locally on dedicated Nvidia GeForce RTXTM PCs and laptops or Nvidia RTXTM-powered workstations, as well as Nvidia DGX Station and Nvidia DGX Spark supercomputers from Altos Computing, ASUS, Dell Technologies, GIGABYTE, HP, Lenovo, and Supermicro.
Enterprises can also build and run AI agents on AI factory infrastructure from cloud service providers including Amazon Web Services (AWS), Google Cloud, Microsoft Azure, Microsoft Security and Oracle Cloud Infrastructure, and on servers from Cisco, Dell Technologies, HPE, Lenovo and Supermicro.