Nutanix Bets Big On Agentic AI With New Controls For The Hybrid Cloud

Nutanix is extending its cloud platform for production agentic AI with new inference, governance, Kubernetes and multitenancy capabilities aimed at helping enterprises and service providers build, run and secure AI agents across hybrid environments.

Nutanix has enhanced and expanded its Nutanix Cloud Platform for production agentic AI with the addition of a dual-native architecture featuring a new Model Context Protocol, or MPC, agent gateway as well as streamlined container management.

San Jose, Calif.-based Nutanix Wednesday unveiled Nutanix Enterprise AI 2.8, which offers centralized control for AI inference and agentic AI.

Also new is Nutanix Kubernetes Platform 2.19, which is slated to streamline container management for bare-metal and virtualized environments. NKP 2.19 includes a built-in AI catalog for building and running agentic AI applications.

Thomas Cornely, Nutanix’s executive vice president of product management, told CRN the enhancements come at a time when agents are running on CPUs and GPUs and containers on both legacy and virtual machine infrastructure.

“You have your core infrastructure, your agentic, compute-centric tier, and your intelligence tier,” Cornely said. “Very few companies, and I would argue actually no other companies, have a platform that allows you to support all of this and give you a consistent way to govern, monitor, operate and just build these end-to-end solutions. This is where Nutanix comes to play.”

Cornely said Nutanix is combining the core pieces enterprises need to build, run, secure and govern agentic AI across hybrid environments, giving IT a consistent way to control how agents access cloud models and enterprise applications and data on-premises or in public clouds.

That work builds upon 16 years of Nutanix platform development and applies it to an AI model in which agents run on containers, consume applications on virtual machines, and connect to GPU-based intelligence in the cloud or on-premises.

Nutanix Enterprise AI 2.8 gives customers a centralized agent gateway for AI inference and agentic AI, Cornely said.

Customers are already seeing token consumption and costs rise as AI use expands, making visibility and controls critical, he said. The gateway lets customers see who is consuming tokens, which models they are using and how much they are spending. IT teams can then set access policies, cap usage, and route workloads to the right models based on cost and performance.

“Not all tasks should be getting tokens from the most expensive, highest-performance model,” he said. “Use some of the high-end models for the most advanced requests. Use your open-weight models for your more common requests.”

NAI 2.8 also includes private inferencing and a centralized MCP layer to govern how agents access applications. Cornely said the goal is to avoid fragmented MCP configurations and give IT one place to manage policies for developers and AI builders.

The next layer is Nutanix Kubernetes Platform 2.19, which underpins NAI because agentic AI workloads run on containers, Cornely said. NKP originally came from Nutanix’s D2IQ acquisition, and is Cloud Native Computing Foundation-compliant, open-source centric, and designed to simplify Kubernetes deployment, management and multitenant operations.

NKP demand is rising as more agents run on containers, Cornely said. NKP 2.19 adds Cloud Native Computing Foundation AI conformance, GPU optimization and an AI catalog with open-source components plus Nutanix’s AI Gateway, private inferencing and MCP capabilities.

“We’re also adding into NKP 2.19 an AI catalog which basically will provide a set of services built into the platform to make it easier for AI builders to build agents using [NKP],” he said. “They’re complemented by some of our own, like our AI Gateway, our NAI for private inferencing, our MCP servers.”

Cornely said NKP 2.19’s “dual-native” architecture can run on Nutanix’s AHV hypervisor, bare metal or public clouds, giving customers flexibility to test AI services in the cloud or on available on-premises infrastructure before moving into production.

That architecture also supports Nutanix’s service provider strategy. Cornely said Service Provider Central, or SP Central, extends the management plane with service provider-governed multitenancy across virtual machines, data, networks and containers, helping MSPs modernize VMware-based environments while adding AI services.

“SP Central allows you to do VMs and containers and do more advanced AI services at the tenant level,” he said. “This is good for MSPs because they’re all modernizing and extending their set of offerings.”

Cornely said neoclouds that now serve a small number of large tenants will need more agile multitenancy as they target enterprise customers. Nutanix sees VMs, containers, AI services and SP Central as the foundation for that shift.

Anthony Jackman, chief innovation officer of Pittsburgh, Pa.-based solution provider, data center services provider and Nutanix channel partner Expedient, said Expedient’s entire AI product line is cloud-native by design, running on Kubernetes, and its default is NKP.

“Every customer of ours gets their own NKP cluster where we run all of the services we’re providing to them,” Jackman told CRN. “But we’re also working with Nutanix’s gateway product and NAI. Our offering is not quite available to the market yet but will be within a month. We’ve been running it in the lab and working with Nutanix’s development team for many months.”

Nutanix's history is a continuing attempt to make everything as easy and consumable as possible, and that's a very positive thing, Jackman said.

“It’s not always as relevant for our clients directly because they have us there to do it for them,” he said. “It does benefit us in that we can make it easier to operate at scale, make it more consistent across clients, which leads to a lower price and a better experience.”

Jackman said Expedient has been working with Nutanix on multitenancy for several years and was the first customer to use it.

“We worked with their engineering team, and we’re going to have our offering out the door shortly to customers, which is I think really going to open up the number of customers that can take advantage of this,” he said. “It’s the quickest, most cost-effective way to consume cloud, and it’s largely the same Nutanix experience that we’ve always known. It just opens it up to more clients to start small and get big.”

Jackman also said NAI 2.8 and its new MCP gateway is very important to customers.

“You know AI that doesn’t have hands doesn’t really do much for you,” he said. “And giving it hands that are not controlled is something that enterprises should not do. This is about centralizing control and making it easy. We’re leveraging it to put a security wrap around AI and make it easier for our clients to adopt it in a secure manner. I think it’s really great to see them expanding that capability. They’re listening to their customers and listening to the market.”