NVIDIA Agent Toolkit at GTC 2026: 17 software giants join the enterprise AI agent platform

Summary

Jensen Huang took the GTC 2026 stage in San Jose carrying what VentureBeat describes as "the blueprints for a new kind of monopoly." The NVIDIA CEO unveiled the Agent Toolkit — an open-source platform for building autonomous AI agents — and immediately announced that 17 major enterprise software companies would build their next generation of AI products on it: Adobe, Salesforce, SAP, ServiceNow, Siemens, CrowdStrike, Atlassian, Cadence, Synopsys, IQVIA, Palantir, Box, Cohesity, Dassault Systèmes, Red Hat, Cisco, and Amdocs.

The toolkit provides the models (the open-source Nemotron family), a secure runtime (OpenShell), an orchestration framework (AI-Q Blueprint), and an optimization library (cuOpt) — everything an AI agent needs to operate autonomously inside an organization, from resolving customer service tickets to designing semiconductors or managing clinical trials. Every component is open source. Every component is optimized for NVIDIA hardware. The combination means that as AI agents proliferate across the corporate world, they will generate demand for NVIDIA GPUs not because companies choose to buy them, but because the software they depend on was engineered to require them.

"The enterprise software industry will evolve into specialized agentic platforms," Huang told the crowd, "and the IT industry is on the brink of its next great expansion." What he left unsaid, according to VentureBeat, is that NVIDIA has just positioned itself as the tollbooth at the entrance to that expansion — open to all, owned by one.

In practice

The Agent Toolkit consists of four core components:

• **Nemotron**: a family of open-source models optimized for agentic reasoning.
• **AI-Q Blueprint**: a hybrid architecture where frontier models handle high-level orchestration while the lighter Nemotron models handle research and retrieval tasks. NVIDIA says this approach cuts cost per query by more than 50% without sacrificing accuracy. The company claims its AI-Q-powered agent currently tops both the DeepResearch Bench and DeepResearch Bench II leaderboards.
• **OpenShell**: an open-source runtime that creates isolated sandboxes enforcing strict policies around data access, network reach, and privacy boundaries. NVIDIA is collaborating with Cisco, CrowdStrike, Google, Microsoft Security, and TrendAI to integrate OpenShell with their existing security tools.
• **cuOpt**: a skill library for optimization tasks.

Partner integrations reveal concrete, large-scale use cases:

• **Salesforce**: uses the Toolkit with Nemotron models so customers can build and deploy Agentforce agents for service, sales, and marketing. The collaboration introduces a reference architecture where Slack serves as the primary conversational interface and orchestration layer — agents participate directly in business workflows and pull from both on-premises and cloud data stores. For the millions of knowledge workers who already conduct their professional lives inside Slack, this turns a messaging app into the command center for corporate AI.
• **SAP**: uses the Toolkit and NeMo to enable AI agents through Joule Studio on SAP Business Technology Platform, allowing customers and partners to design agents tailored to their specific business needs.
• **Adobe**: will adopt the Toolkit as the foundation for hybrid, long-running agents across creativity, productivity, and marketing. CEO Shantanu Narayen stated the partnership brings together Firefly models, CUDA libraries, 3D digital twins for marketing, and the Agent Toolkit to "deliver high-quality, controllable and enterprise-grade AI workflows of the future." Adobe will also explore OpenShell and Nemotron for personalized, secure agentic loops and evaluate the toolkit for large-scale workflows powered by Adobe Experience Platform.
• **ServiceNow**: its Autonomous Workforce of AI Specialists leverages the Toolkit, the AI-Q Blueprint, and a combination of closed and open models — including Nemotron and ServiceNow's own Apriel models — suggesting the toolkit is designed not to replace existing AI investments but to become the connective tissue between them.
• **Siemens**: is launching the Fuse EDA AI Agent, which uses Nemotron to autonomously orchestrate workflows across its entire electronic design automation (EDA) portfolio, from design conception through manufacturing sign-off, for the semiconductor and PCB industries.
• **Cadence**: integrates the Toolkit and Nemotron into the ChipStack AI SuperAgent for semiconductor design and verification.
• **Synopsys**: is building a multi-agent framework using its AgentEngineer technology with Nemotron and the NeMo Agent Toolkit.
• **IQVIA**: integrates Nemotron and other Toolkit components with IQVIA.ai, a unified agentic AI platform for life sciences organizations. The scale is already significant: IQVIA has deployed more than 150 agents across internal teams and client environments, including 19 of the top 20 pharmaceutical companies.
• **CrowdStrike**: is launching a Secure-by-Design AI Blueprint that embeds its Falcon platform protection directly into NVIDIA AI agent architectures, including agents built on AI-Q and OpenShell.
• **Palantir**: is developing AI agents on Nemotron running on its sovereign AI Operating System Reference Architecture.
• **Atlassian**: integrates the Toolkit into its Rovo AI agentic strategy for Jira and Confluence.
• **Box**: uses the Toolkit to enable enterprise agents to securely execute long-running business processes using the Box file system.

Deployments are supported across major cloud providers: AWS, Google Cloud, Microsoft Azure, and Oracle Cloud Infrastructure. Developers can explore the Agent Toolkit and OpenShell on build.nvidia.com, running on inference providers and NVIDIA Cloud Partners including Baseten, CoreWeave, DeepInfra, and DigitalOcean — all running NVIDIA GPUs.

NVIDIA also announced the **Nemotron Coalition**: a global collaboration including Mistral AI, Cursor, LangChain, Perplexity, Reflection AI, Sarvam, and Thinking Machines Lab, aimed at advancing open frontier models. The coalition's first project will be a base model co-developed by Mistral AI and NVIDIA, trained on NVIDIA DGX Cloud, which will underpin the upcoming Nemotron 4 family. LangChain — whose open-source agent frameworks have been downloaded over one billion times — will integrate AI-Q, OpenShell, and the Nemotron model family into its deep agent library.

What we still don't know

VentureBeat flags several risks and limitations that enterprise buyers should weigh before going all-in on this platform.

First, **adoption announcements are not deployment announcements**. Many partner disclosures use carefully hedged language — "exploring," "evaluating," "working with" — standard in embargoed press releases but not to be confused with production systems serving millions of users. Adobe's own forward-looking statements note that "due to the non-binding nature of the agreement, there are no assurances that Adobe will successfully negotiate and execute definitive documentation with Nvidia on favorable terms or at all."

Second, **NVIDIA is not the only company chasing this market**. Microsoft, with its Copilot ecosystem and Azure AI infrastructure, pursues a parallel strategy with the advantage of owning the operating systems and productivity software most enterprises already use. Google, through Gemini and its cloud platform, has its own agent vision. Amazon, via Bedrock and AWS, is building comparable primitives. The question is not whether enterprise AI agents will be built on some platform, but whether the market will consolidate around one stack or fragment across several.

Third, **the security claims, while architecturally sound, remain unproven at scale**. OpenShell's policy-based guardrails are a promising design pattern, but autonomous agents operating in complex enterprise environments will inevitably encounter edge cases that no policy framework has anticipated. CrowdStrike's Secure-by-Design AI Blueprint and Cisco AI Defense's OpenShell integration are exactly the kind of layered defense enterprise buyers will demand — but both are newly unveiled, not battle-hardened through years of adversarial testing.

Finally, there is the question of whether **enterprises are ready for agents at all**. The technology may be available, but organizational readiness — governance structures, change management, regulatory frameworks, human trust — often lags years behind what the platforms can deliver.

Why it matters

• The simultaneous adoption by 17 software companies touching virtually every industry and every Fortune 500 corporation represents an unprecedented collective endorsement of NVIDIA as the default substrate for enterprise intelligence — and a clear signal that the market sees autonomous agents as the next inevitable phase of corporate work.
• The Android analogy used by VentureBeat to describe the strategy reveals the underlying platform logic: by distributing software optimized for its GPUs at no cost, NVIDIA creates structural dependency across the ecosystem without relying on contractual exclusivity.
• The Salesforce/Slack integration has the potential for direct and immediate impact on millions of knowledge workers, turning a communication tool into a central control point for corporate AI.
• The 50%-plus reduction in cost per query claimed by NVIDIA — if independently validated — makes the economic case for adopting autonomous agents significantly more accessible, removing one of the main barriers to large-scale enterprise adoption.