Tenant Clusters for Production Kubernetes and AI Infrastructure
Virtual control planes, real isolation — from a single node to 100K-GPU superclusters.
CNCF Certified Kubernetes — Distribution · Kubernetes AI Conformant
vCluster Labs builds the open-source foundation for running production Kubernetes and AI factories on your own hardware. Our flagship product, vCluster, creates Tenant Clusters — fully isolated Kubernetes environments with their own API server, CRDs, and RBAC, indistinguishable from a dedicated cluster.
Built for production. Trusted in production. 40M+ Tenant Clusters deployed by teams at Adobe, CoreWeave, NVIDIA, Lintasarta, Atlan, Deloitte, and hundreds of AI clouds, AI factories, and Fortune 500 platform organizations.
The public-cloud experience, on your own infrastructure. Give every team the Kubernetes they need — with strict tenant isolation, hardware-aware scheduling, and zero tenant sprawl — whether you run one region or 100K GPUs.
| Project | What it does |
|---|---|
| vCluster | Tenant Clusters for Kubernetes — Virtual Control Planes with API, data, and (optionally) network isolation. |
| vind | vCluster in Docker — like kind, but with the full vCluster feature set (UI, sleep/resume, LoadBalancer, image cache). |
| Cluster API Provider vCluster | Manage vCluster Tenant Clusters with the upstream Cluster API. |
| vCluster SDK | Build custom vCluster syncers and extend Tenant Cluster behavior. |
| vCluster Docs | Documentation for vCluster, the Platform, and the surrounding ecosystem. |
| Certified Stacks | Curated, validated Kubernetes stacks (CNI, CSI, GPU, observability) that ship as the foundation of an AI factory. |
| vCluster Platform | The control plane for self-service Kubernetes — manage Tenant Clusters, namespaces, and access at scale. |
| DevPod | Codespaces, but open-source and unopinionated — works with any IDE and any cloud, Kubernetes, or Docker. |
| jsPolicy | Easier and faster Kubernetes policies using JavaScript or TypeScript. |
vCluster is the foundation of a broader platform for running production Kubernetes and AI infrastructure on your own hardware — from a single rack to 100K-GPU supercomputers.
| Layer | Project | Purpose |
|---|---|---|
| Tenant Clusters | vCluster | Virtual Control Planes with API, data, and (optionally) network isolation |
| Runtime Isolation | vNode | Kernel-enforced workload sandboxing (seccomp, cgroups, namespaces, AppArmor) — no VM overhead |
| GPU Infrastructure | vMetal | Zero-touch bare-metal provisioning for GPU fleets |
| Network Isolation | Netris (integration) | Hardware-enforced isolation via programmatic VLANs, VRFs, and ACLs |
Together these provide a complete foundation for AI factories — certified Kubernetes stacks, isolated tenant clusters, runtime workload sandboxing, and GPU infrastructure operations — the same pattern used to run production AI on hundreds of AI clouds and Fortune 500 on-prem platforms.
- AI Factory — Run AI on-prem where your data and GPUs live
- AI Cloud Providers — A hyperscaler-like Kubernetes experience for GPU customers
- Internal GPU Platform — Maximize GPU utilization without sacrificing isolation
- Bare Metal Kubernetes — Production Kubernetes on bare metal, no VMs
- Software Vendors — Ship Kubernetes-native products with per-customer Tenant Clusters
- Cost Savings — Consolidate clusters, sleep idle workloads, cut Kubernetes spend
Real usage, not a gated demo. Unlimited Tenant Clusters up to 64 CPUs / 32 GPUs, plus the full vCluster Platform UI — for free.
Get Started Free → • Quickstart • Try in your browser →
- 💬 Join our Slack community — 5K+ engineers
- 💼 Follow on LinkedIn — 28K+
- 🐦 Follow on X — 3.7K+
- 📺 Watch on YouTube
- 📝 Read the Blog
- 📅 Book a consultation
⭐ Star our projects if they help you build better Kubernetes infrastructure!
