Nscale Kubernetes Service (NKS)
Deploy and scale managed Kubernetes on Nscale GPU infrastructure to build, train, and deploy AI applications.
The Kubernetes you know, without the admin burden
Your teams want Kubernetes as the workload interface, without taking on the operational burden of the control plane.
NKS keeps your workflows familiar: create clusters and worker nodes through the NKS API, and keep the Kubernetes APIs, manifests, and RBAC your teams already use.
Nscale operates the control plane and cluster lifecycle, so your platform team can focus on running applications rather than managing Kubernetes infrastructure.
Why use the Nscale Kubernetes Service
Create clusters through the NKS API and keep the standard Kubernetes tooling your team already uses so workloads stay portable.
NKS combines enterprise-grade clusters with GPU worker pools, topology-aware reservation and placement, networking, industry-standard storage protocols, and load balancing, to meet internal demand.
Nscale handles control-plane availability, upgrades, and monitoring, so your platform team spends more time on AI workloads and less on Kubernetes administration.
How it works
Request a cluster
Submit a cluster request to the NKS API, referencing a network in your Nscale project.
Cluster provisioning and operation
NKS validates your request, optimizes placement for performance, provisions worker nodes, and creates and operates the control plane.
Run your workloads
Use Kubernetes APIs for workloads and permissions. Node pools have their own lifecycle, so you can add and scale worker capacity through NKS without recreating the cluster.
Built for AI teams
NKS runs inside Nscale Cloud, connecting managed Kubernetes to the identity, networking, and compute underneath it.
Authentication uses your Nscale identity, mapped into standard Kubernetes users and project-scoped groups so that Kubernetes RBAC decides what each identity can do inside the cluster. Clusters can be configured with a public Kubernetes API server, or a private API server reachable only from inside your VPC.
For AI Natives
Run training, inference, and supporting services with the Kubernetes APIs, manifests, RBAC, and tooling your team already knows.
For enterprises
Support enterprise requirements for governance, identity, isolation, and network access with managed Kubernetes that fits your cloud operating model.
For developers
Build and ship applications with familiar Kubernetes tools and development workflows, while Nscale operates the clusters underneath.