Introduction

Longbridge Control Plane is a managed Kubernetes control plane. We run the part of a cluster that is genuinely annoying to operate well — the API server, etcd, certificates, and upgrades — and you attach your own machines as nodes, wherever they are.

You get a standard, conformant cluster and a plain kubeconfig. You never get a bill that grows with the hardware you already own.

In private beta. The Control Plane is available to early-access users while we harden it. Commands and defaults in these docs can still change; we note where that’s likely.

The one idea

A Kubernetes cluster has two halves:

• The control plane — the brain. It schedules work, stores cluster state, and exposes the API. It’s fiddly to run reliably and it doesn’t run your apps.
• The nodes — the muscle. This is the compute that actually costs money, and it’s where your workloads run.

Most managed Kubernetes charges you for the control plane and taxes every node you bring. Longbridge runs the control plane for a flat price and leaves the nodes entirely to you — a NUC under your desk, a Hetzner box, an AWS spot instance, joined as one cluster.

Who it’s for

• Homelabbers and self-hosters who want a real, managed control plane without paying per node for machines they own.
• Small platform teams running production across more than one place — on-prem, a cheap European cloud, and a hyperscaler — who want one cluster spanning all of it.

It is not a place to rent compute. You bring the compute; we run the brain.

What you get

• A managed, patched, monitored control plane in a region you choose.
• Bring-your-own nodes across any cloud or your own hardware — always free.
• One kubeconfig and standard kubectl. Nothing proprietary to learn.
• Flat, honest pricing by control-plane size, starting at €7/mo — never per node, per vCPU, or per gigabyte you own.

Where to next

• New here? The Quickstart gets you a cluster and a joined node in a few minutes.
• Want the mental model first? Read Core concepts.