vCluster Docs

vCluster powers highly scalable, resilient compute clusters and gives AI cloud providers a self-service platform for offering them to tenants. Deliver managed Kubernetes, Slurm, inference, and other cluster products with tenant management, capacity management, multi-tenant observability, access control, and lifecycle automation built in. Add vMetal to expose physical and virtual infrastructure through a stable Machine API backed by pluggable compute and network drivers.

Search...

Where do you want to start?

rocket_launch

Prove vCluster on your infrastructure

Start with the hardware you have or Docker and launch your first managed compute cluster.route

Find your production path

Match your use case, inference, AI cloud, CI/CD, and more, to a concrete build and operate path.dns

Provision your infrastructure

Turn bare metal or virtual machines, GPUs included, into tenant cluster capacity.hub

Offer self-service clusters

Manage tenants, capacity, templates, access, and multi-tenant observability from one platform.

Not sure where to start? See how the products fit together to understand how they divide responsibility, then pick the entry point that matches your task.

How the products fit together

vCluster

Managed Kubernetes, Slurm, inference, and other compute clusters with hyperscaler-grade isolation and reliability. What is vCluster?

vCluster Platform

Provides self-service tenant management, capacity management, templates, multi-tenant observability, policy, and access control. What is vCluster Platform?

vMetal

Exposes physical and virtual infrastructure through a stable Machine API backed by interchangeable provisioning and network drivers. vMetal overview

Tenant clusters and vCluster Platform combine to form vCluster. vMetal connects vCluster to physical and virtual infrastructure through a stable Machine API.

See how the products fit together for the full request-to-infrastructure workflow, the boundaries between each product, and where to go next.

Try vCluster Platform

Create your own platform instance and offer self-service compute clusters to tenants. Manage tenants, capacity, access, and operations across all your environments.

Get Started →