As previously posted on the Nutanix Tech Center site, we discussed integrating Nutanix on Google Cloud with the Gemini Enterprise Agent Platform (formerly known as the Vertex AI platform). The blog showcases how easy it is to connect to native public cloud services from workloads running in Nutanix Cloud Clusters (NC2) running within the hyperscaler of choice.
Although the blog and the video below depict a very straightforward example demo, the possibilities are vast! The Nutanix Platform delivers the freedom and flexibility to use infrastructure services while keeping operational consistency across the board. The value lies in combining all the provided platform services, with workloads running in a hybrid fashion across private and public clouds.
In the provided example, we have an inference workload running in Nutanix clusters, hooking into Google-native services. By the way, Google Cloud was used as an example. We can do the same with services like AWS Bedrock or Microsoft Azure AI Foundry. To show you what good, better, and best look like, let’s dive a bit deeper into how the Nutanix Cloud Platform further complements the setup used in the blog and demo!
Good
Workloads were migrated from on-prem data centers into public cloud using Nutanix Cloud Clusters (NC2). The dataset resides within the Nutanix platform, but the compute power to train your model of choice using the Gemini Enterprise Agent Platform and perform inference in Nutanix clusters. This is what is shown in the blog and in the demo linked below. A great first step: freedom to run your workloads wherever you want, and optionally extend with native cloud capabilities!
Better
To maintain better control and governance over used AI API endpoints, such as Gemini in this example, front this with the newly announced Nutanix Agent Gateway. This is a great way to have one control point for connecting agents and applications to public, private, and self-hosted models, while enforcing policy, providing observability, and controlling token costs. Think about the ability to limit token spend per model, or large-language-model failovers in the event of AI API endpoint outages.
Unlike a standalone gateway, it integrates with Nutanix Enterprise AI so customers can choose the right model, infrastructure, and deployment location without cloud or vendor lock-in.
Best
Nutanix Enterprise AI (NAI) offers endpoint APIs for leading LLM providers, including NVIDIA NIM and HuggingFace, making it easy for organizations to deploy a wide range of Gen AI models on-premises or in the public cloud. NAI includes a simple UI, role-based access controls (RBAC), and the capability for untethered operations (dark site or air-gapped deployments) to simplify the operation, monitoring, and adaptation of AI models (LLMs) with enterprise resiliency, day 2 operations, and compliance. Combined with the Nutanix Agent Gateway, it allows customers to choose the right model, infrastructure, and deployment location without cloud or vendor lock-in.
To run NAI, you’ll need Kubernetes (K8S). In comes our Nutanix Kubernetes Platform (NKP) to efficiently run and operate your K8S estate. Oh, you need vector databases as well? Use Nutanix Database Service (NDB) to offload all database lifecycle operations. The Nutanix Platform provides all is the key takeway here.
See for yourself
This is a video detailing the demo, as recorded at the Nutanix .NEXT 2026 event in Chicago! It focuses on the lightest-weight approach: connecting workloads in Nutanix to the Gemini Enterprise Agent Platform.
