Both are free, open-source Kubernetes UIs with desktop apps. Freelens actively carries the familiar Lens model forward through community maintenance and plugins. Radar starts from a newer premise: resources, relationships, changes, Issues, GitOps, Helm, traffic, and operator state should work as one live model.
Both let you inspect and edit resources. Radar starts where resource browsing stops.
The overlap is real. The difference is where each product puts the complexity.
Resource browsing and editing, YAML, logs, exec, port-forwarding, cluster metrics when available, kubeconfig access, and account-free local use are not the wedge. The decision starts after that baseline.
Radar's depth is not the number of tabs. It is the context that follows you between them.
Radar Issues ranks what is failing now and carries the context needed to triage it: missing references, scheduling blockers, recent changes, and links to affected resources. The topology is not a separate picture - it is another view of the same relationships.
Read how Issues worksThe timeline merges Kubernetes events, resource diffs, pod lifecycle, GitOps activity, and Helm operations. It runs in memory by default, persists locally with SQLite, and adds cloud-backed retention up to one year with Radar Cloud.
Read the timeline docsRadar connects sync state, drift, managed resources, operation history, controller health, and actions. The same ownership then appears in resource drawers, topology, timeline, Helm, Issues, and MCP instead of living inside a standalone plugin panel.
Explore GitOps diagnosisFreelens has real, actively maintained extensions. Radar makes a different tradeoff: typed integrations for Argo CD, Flux, Karpenter, KEDA, cert-manager, Kyverno, Trivy, OpenCost, Prometheus, Gateway API, and more ship in the binary and activate when their CRDs are present and readable.
That gives topology, resource drawers, Issues, actions, and MCP the same understanding of each controller, on one version and one update cycle.
See every integrationOfficial Linux amd64 release downloads measured August 4, 2026.
Radar modes are labeled explicitly. Freelens capabilities distinguish core, official extensions, and community extensions.
| Feature | Radar Desktop · local · in-cluster · Cloud | Freelens Desktop IDE · extensions |
|---|---|---|
| Product and delivery | ||
License and governance Both licenses preserve the rights granted by existing releases. The governance models differ: vendor-led commercial OSS versus an independent community project. | Apache 2.0 · Skyhook-led · no-relicense pledge | MIT · community-led |
Desktop app Both ship desktop applications for macOS, Linux, and Windows. Radar uses the operating system webview around its compiled Go core; Freelens bundles Electron, Chromium, and Node. | Strong: macOS · Linux · Windows | Strong: macOS · Linux · Windows |
Local browser mode Radar can run as a CLI or kubectl plugin and serve the same UI on localhost. Freelens is a desktop application. | Strong: CLI · kubectl plugin · browser UI | No |
Account required for local use | Strong: No | Strong: No |
Everyday cluster management Resource browsing and editing, YAML, logs, exec, port-forwarding, and metrics. | Strong: Built in | Strong: Built in |
| Operational context | ||
Topology graph Freelens has an official Resource Map extension. Radar topology is part of the core model and connects to Issues, traffic, GitOps, audit findings, resource drawers, and MCP. | Strong: Live workload graph | Partial: Official extension |
What is broken right now? Radar Issues ranks active failures, missing references, scheduling blockers, and failing operator conditions, then correlates related resources and recent changes. | Strong: Ranked Issues stream in core | Partial: Events, logs, metrics, and extensions |
Retained event and resource-change history Cluster events and resource changes the product itself retains for look-back - not live API state, not your monitoring stack's storage, and not a log of user actions in the product. Freelens reads live Kubernetes events and charts metrics by querying a Prometheus or metrics-server you run in the cluster; it retains no event or resource-change history of its own. | Strong: Memory · SQLite · Cloud up to 1 year | No |
Resource-change diffs Field-level diffs of what actually changed on a resource, not just that an event fired. Freelens shows current YAML and Kubernetes events; it does not record prior revisions to diff against. | Strong: Timeline + manifest history | No |
GitOps correlation (Argo CD + Flux) Generic CRD access is useful; typed sync status, drift, ownership, activity, and actions wired into the rest of the product is a deeper integration. Radar treats Argo CD and Flux as one core workspace with topology, drift, activity, diagnosis, and actions. Freelens has an official Flux extension and a maintained community Argo CD extension. | Strong: Argo CD + Flux, in core | Partial: Official + community extensions |
Helm diagnosis and management Radar reads Helm release secrets directly from Kubernetes, compares revisions, diagnoses failed operations and hooks, and supports upgrade, rollback, and uninstall without requiring the Helm CLI. | Strong: Core · no Helm CLI | Strong: Bundled Helm workflows |
Kubernetes ecosystem integrations Radar compiles typed renderers into one release and activates them when their CRDs are discovered and readable. Freelens combines generic CRD access with official and community extensions. | Strong: Typed · auto-detected · one release | Partial: Extensions + generic CRDs |
MCP for external AI agents What an external agent actually gets from the endpoint: a wrapper over the raw Kubernetes API, or operational context already correlated for agent use. Benchmarked against raw kubectl on 54 live-cluster faults, Radar's MCP server was more accurate and reached a correct diagnosis about four times sooner. Radar exposes that model to external MCP clients. Freelens offers an official in-app AI extension that acts as an MCP client. | Strong: Correlated read + write tools · 4x sooner to a correct diagnosis than kubectl | Partial: In-app AI client extension |
Live workload-to-workload traffic + mTLS Radar maps observed workload-to-workload traffic and service-mesh encryption in the same topology. No equivalent feature is listed in the official Freelens extension list. | Yes | No |
Cluster security audit (Trivy / Kyverno) Configuration, reliability, and security findings surfaced on the affected resources - and whether that ships with the product or arrives as a separately installed, separately licensed scanner. Radar's built-in checks surface those findings on the affected resources, with typed Trivy and Kyverno context when those controllers are present. | Yes | Partial: Generic resources + extensions |
| Teams and fleets | ||
Shareable team URL A local Radar session is local. Deploy Radar in-cluster for a shared URL to one cluster, or use Cloud. Freelens runs on each engineer's laptop. | Strong: In-cluster or Cloud | Per-laptop |
Audit log retention A record of who did what in the product, distinct from the cluster event timeline. Radar Cloud retention is 7 days on Free, 30 days on Team, and 365 days on Enterprise. | Strong: Cloud · 7 / 30 / 365 days | No |
Slack and webhook alerts Routing Kubernetes state changes out of the product to where the team already works. | Strong: Cloud | No |
Hosted SaaS option | Strong: Cloud | No |
Clusters in one view Radar desktop and local modes switch kubeconfig contexts but operate on one active cluster; Radar Cloud aggregates across connected clusters. Freelens catalogs multiple kubeconfig clusters and opens one at a time. | Strong: Cloud scoped fleet views | Partial: Catalog · one open at a time |
Cross-cluster operations The row above asks whether you can see several clusters at once; this one asks what actually aggregates across them. Radar Cloud aggregates Issues, applications, search, checks, and package drift. Cross-cluster topology and timeline are not available today. | Strong: Cloud · Issues, search, checks, drift | No |
| Auth | ||
Organization SSO Product sign-in for shared or team modes. Freelens runs on each engineer's laptop against their own kubeconfig, so it has no product sign-in to federate - the official extension list has no SSO or team-authentication extension either. | Strong: In-cluster OIDC · Cloud SAML/OIDC | No |
SCIM provisioning | Partial: Enterprise | No |
Inherits Kubernetes RBAC Both reach the cluster as the user, so the Kubernetes API enforces the same permissions either tool would get from kubectl. | Yes | Yes |
| Architecture and footprint | ||
Runtime model Radar's compiled Go core uses the official Kubernetes client libraries and informer/watch primitives - the same machinery used by Kubernetes controllers and operators - plus Go's lightweight concurrency to process independent cluster work in parallel. It pushes deltas over SSE and uses the operating system webview on desktop; Freelens packages Electron, Chromium, and Node. | Strong: Compiled Go + system webview | Electron + Chromium + Node |
Linux amd64 .deb download Current official releases measured August 4, 2026. Radar is about one quarter the size of Freelens. | Strong: ~40 MB | ~150 MB |
Both licenses preserve the rights granted by existing releases. The governance models differ: vendor-led commercial OSS versus an independent community project.
Both ship desktop applications for macOS, Linux, and Windows. Radar uses the operating system webview around its compiled Go core; Freelens bundles Electron, Chromium, and Node.
Radar can run as a CLI or kubectl plugin and serve the same UI on localhost. Freelens is a desktop application.
Resource browsing and editing, YAML, logs, exec, port-forwarding, and metrics.
Freelens has an official Resource Map extension. Radar topology is part of the core model and connects to Issues, traffic, GitOps, audit findings, resource drawers, and MCP.
Radar Issues ranks active failures, missing references, scheduling blockers, and failing operator conditions, then correlates related resources and recent changes.
Cluster events and resource changes the product itself retains for look-back - not live API state, not your monitoring stack's storage, and not a log of user actions in the product. Freelens reads live Kubernetes events and charts metrics by querying a Prometheus or metrics-server you run in the cluster; it retains no event or resource-change history of its own.
Field-level diffs of what actually changed on a resource, not just that an event fired. Freelens shows current YAML and Kubernetes events; it does not record prior revisions to diff against.
Generic CRD access is useful; typed sync status, drift, ownership, activity, and actions wired into the rest of the product is a deeper integration. Radar treats Argo CD and Flux as one core workspace with topology, drift, activity, diagnosis, and actions. Freelens has an official Flux extension and a maintained community Argo CD extension.
Radar reads Helm release secrets directly from Kubernetes, compares revisions, diagnoses failed operations and hooks, and supports upgrade, rollback, and uninstall without requiring the Helm CLI.
Radar compiles typed renderers into one release and activates them when their CRDs are discovered and readable. Freelens combines generic CRD access with official and community extensions.
What an external agent actually gets from the endpoint: a wrapper over the raw Kubernetes API, or operational context already correlated for agent use. Benchmarked against raw kubectl on 54 live-cluster faults, Radar's MCP server was more accurate and reached a correct diagnosis about four times sooner. Radar exposes that model to external MCP clients. Freelens offers an official in-app AI extension that acts as an MCP client.
Radar maps observed workload-to-workload traffic and service-mesh encryption in the same topology. No equivalent feature is listed in the official Freelens extension list.
Configuration, reliability, and security findings surfaced on the affected resources - and whether that ships with the product or arrives as a separately installed, separately licensed scanner. Radar's built-in checks surface those findings on the affected resources, with typed Trivy and Kyverno context when those controllers are present.
A local Radar session is local. Deploy Radar in-cluster for a shared URL to one cluster, or use Cloud. Freelens runs on each engineer's laptop.
A record of who did what in the product, distinct from the cluster event timeline. Radar Cloud retention is 7 days on Free, 30 days on Team, and 365 days on Enterprise.
Routing Kubernetes state changes out of the product to where the team already works.
Radar desktop and local modes switch kubeconfig contexts but operate on one active cluster; Radar Cloud aggregates across connected clusters. Freelens catalogs multiple kubeconfig clusters and opens one at a time.
The row above asks whether you can see several clusters at once; this one asks what actually aggregates across them. Radar Cloud aggregates Issues, applications, search, checks, and package drift. Cross-cluster topology and timeline are not available today.
Product sign-in for shared or team modes. Freelens runs on each engineer's laptop against their own kubeconfig, so it has no product sign-in to federate - the official extension list has no SSO or team-authentication extension either.
Both reach the cluster as the user, so the Kubernetes API enforces the same permissions either tool would get from kubectl.
Radar's compiled Go core uses the official Kubernetes client libraries and informer/watch primitives - the same machinery used by Kubernetes controllers and operators - plus Go's lightweight concurrency to process independent cluster work in parallel. It pushes deltas over SSE and uses the operating system webview on desktop; Freelens packages Electron, Chromium, and Node.
Current official releases measured August 4, 2026. Radar is about one quarter the size of Freelens.
Published · Last updated .
This is where Radar's delivery flexibility matters. Each mode has a specific scope.
One engineer, one active kubeconfig context. No account and nothing installed in the cluster. Switch contexts when you need another cluster.
One shared URL for one cluster, with OIDC or proxy authentication and Kubernetes RBAC enforced per viewer. SQLite persistence is optional.
Aggregate Issues, applications, search, checks, and package drift across connected clusters. Topology and timeline remain per-cluster today.
Use the same interface as a desktop app, local browser, or shared in-cluster service.
$curl -fsSL https://get.radarhq.io | sh && kubectl radarApache 2.0 · No account for local use · Run Radar OSS forever