Radar icon

Radar

Radar is an open-source Kubernetes UI that runs as a single binary on a laptop or in-cluster, with views for topology, events, Helm, GitOps, audits, traffic, and cost data. Radar Cloud adds hosted multi-cluster, retention, alerts, and SSO features on top of the same engine.

Radar

Open-source Kubernetes UI with hosted fleet options

Radar is an open-source Kubernetes UI built around a single Go binary. It runs locally or inside a cluster and is designed to help operators understand workloads, events, topology, GitOps state, cost signals, and traffic without switching between several separate tools.

The open-source engine is Apache 2.0 and requires no account. Radar Cloud uses the same engine as a hosted layer for fleet-scale needs such as multi-cluster aggregation, persistent retention, alerts, and SSO.

Core capabilities

Live topology graph

Visualize resources and their connections as a live topology graph, with grouping by namespace or label and updates pushed over SSE.

Event timeline

Capture Kubernetes events and resource deltas as they happen so teams can review incidents after the API server’s built-in event TTL.

Image filesystem viewer

Inspect container image filesystems directly in the UI, including tree browsing, sizes, symlinks, search, and downloads.

Cluster audit

Run a cluster audit with 31 best-practice checks covering security, reliability, and efficiency, using cached state for fast results.

GitOps and Helm management

Work with GitOps and package data in one place, including FluxCD and ArgoCD sync state, Helm releases, and revision history.

Ecosystem-aware workflows

Add OpenCost cost views, traffic visualization, and MCP tools for AI agents when those integrations are present in the cluster.

Common workflows

  • Incident troubleshooting

    Use Radar when a pod is failing and you need to search across clusters, inspect logs, and rewind the event timeline from one interface instead of hopping between kubectl commands.

  • Multi-cluster operations

    Use the fleet view to compare workloads, search resources across every connected cluster, and keep track of drift or failures in a single signed-in workspace.

  • GitOps visibility

    Use the GitOps views to inspect FluxCD or ArgoCD sync state alongside the resources they produced, especially when you want to understand reconciliation status quickly.

  • Helm release review

    Use the Helm manager to review releases, diff revisions, inspect values files, and roll back changes without leaving the UI.

  • Workload inspection and review

    Use the image filesystem viewer and cluster audit to inspect what is running in a workload and check it against best-practice rules before or after a change.

Pros and Cons

Pros

  • Runs as a single binary in a cluster or on a laptop, which keeps installation and local usage simple.
  • Covers a broad set of operator workflows in one UI, including topology, events, Helm, GitOps, image inspection, audits, traffic, and cost views.
  • Offers a free self-hosted path with Apache 2.0 licensing and no account required.
  • Uses push-based updates and shared-informer-style watching instead of periodic full scans, which the site positions as better for large clusters.
  • Provides a hosted option with per-cluster pricing rather than per-seat pricing, plus a 14-day free trial on paid plans.

Cons

  • The public pages do not spell out detailed self-hosted deployment steps beyond Helm-based install references.
  • Advanced hosted features such as fleet aggregation, longer retention, routed alerts, and SSO are presented as Cloud-only.

FAQ

What is the difference between self-hosting Radar and using Radar Cloud?

Radar OSS runs as a single Go binary you can install with Helm and run in your own cluster, or use locally from your laptop. Radar Cloud layers on hosted fleet features such as multi-cluster view, longer retention, alerts, and SSO.

Can I use Radar OSS and Radar Cloud together?

Yes. The source describes Radar OSS as the open-source engine and Radar Cloud as an optional layer on top. The pages do not describe a combined deployment in detail, so the safest reading is that they are separate ways to use the same engine.

Why is Radar priced per cluster instead of per user?

Pricing is per cluster rather than per seat. The site says this is intended to let teams add viewers without paying more per person.

Is there a free option or trial?

The pricing page states that self-hosted Radar is free forever, while paid hosted plans include a 14-day free trial with no credit card.

Is my cluster data sent to Radar?

The site says cluster data stays in your cluster for the self-hosted engine, while Radar Cloud adds hosted features such as retention and alerts. The public pages do not provide a deeper data-handling policy beyond that.

Quick Facts

Category
Developer Tool
Product type
Open-source Kubernetes UI
Deployment
Local binary or in-cluster Helm install
License
Apache 2.0
Pricing model
Free self-hosted; hosted plans billed per cluster
Website
radarhq.io