Live, not polled
Every list is a watch stream, on by default, so what you see is what is happening in your cluster right now - crashes, rollouts and node pressure appear the instant they occur.
Software Publishing
Your Kubernetes clusters, native and live on macOS.
Kubius
Kubius is a native macOS Kubernetes client. It talks straight to your cluster's API - no kubectl, no Electron, no terminal - with live watch streams for every resource, on by default. What you see is what is happening in your cluster in real time.
It's built in Swift and SwiftUI for macOS 14 and later, for Apple Silicon or Intel. Your cluster data stays between your Mac and your cluster - nothing flows through our servers.
Kubius grew out of our own need for better tooling while running production workloads on Kubernetes, and we now build, license and support it as a product in its own right.
Three things that set it apart
Genuinely native
Real Swift and SwiftUI with native windows and keyboard shortcuts - no kubectl, no Electron, no terminal. It is fast in the way only a native app can be.
Private by default
Kubius talks straight to your cluster API over TLS. Your cluster data stays between your Mac and your cluster - nothing flows through our servers.
What it does
Live watch streams
Lists are backed by Kubernetes watch streams, not polling - pod crashes, rollout progress and node pressure show up the instant they happen.
Multi-cluster, truly native
Open GKE, EKS, AKS, DigitalOcean and self-managed clusters side by side in real SwiftUI windows, each fully isolated.
Zero-setup connection
Reads your existing kubeconfig and speaks the Kubernetes REST API directly over TLS - with support for cloud exec-auth, client certs and bearer tokens.
Live log streaming
Instant load then live-follow, with level detection, regex search and export - plus combined multi-pod logs colour-coded by pod.
Live metrics
CPU and memory charts read straight from metrics.k8s.io, with request, limit and capacity reference lines - no Prometheus to stand up.
Right-sizing recommendations
Samples real usage across replicas, recommends sensible requests and limits, and applies them in place.
Cluster dashboard & diagnostics
A single-screen health grade for the cluster, with node and workload status and diagnostics that flag OOM kills, kubelet flapping and pressure.
RBAC & network policy tools
Resolve effective permissions in both directions, and get plain "can pod A reach pod B?" verdicts that name the responsible network policies.
The full day-to-day toolkit
Apply YAML with server-side validation, port-forward, exec into containers, browse CRDs, drive ArgoCD, and jump anywhere with a ⌘K command palette.
Got a project in mind?
Whether you want to know what we could do for your organisation, a demo of our work, or just a chat - we'd love to hear from you.