Introducing ClusterCost: The Open-Source Kubernetes Cost Tool That’s Actually Simple

We built ClusterCost because cloud cost tools felt like another platform to manage. Engineers wanted a single, fast agent that speaks Kubernetes and stays out of the way. ClusterCost ships cost insights where you work—CLI and manifests—not buried in another dashboard.

What problem it solves

• No-copy pricing: combines your node prices, storage, and egress with in-cluster usage.
• Owner-first reporting: cost per namespace, team, and service using labels you already enforce.
• Budget guardrails: alerts and policy hooks that block cost regressions before deploy.
• Network visibility: egress and cross-AZ traffic surfaced next to compute so surprises stop.
• Audit-ready: price sheets and decisions are versioned in Git; outputs match what finance sees.

Install in minutes

• One Helm chart or npx clustercost@cli install.
• Single Go agent; no extra databases or sidecar sprawl.
• Works on EKS, GKE, AKS, and on-prem with the same manifest.
• Audit mode: install, report, and uninstall automatically for one-off checks.

Why it’s different

• Local-first: data stays in your cluster; export only what you choose.
• Developer-focused: PR comments with cost deltas; CLI summaries, not PDFs.
• Sane defaults: requests/limits checks, waste reporting, and egress visibility on day one.
• Guardrails: admission policies for labels, limits, and budget caps—not just dashboards.
• Footprint: <150MB memory, <100m CPU steady—small enough for every cluster.

Simple tools win because teams actually use them. ClusterCost gives you cost control without another control plane.***