K8sGPT

K8sGPT: AI-Powered Kubernetes Diagnostics

K8sGPT is a CNCF Sandbox open-source tool that uses AI to scan Kubernetes clusters for problems and explain them in plain English. It eliminates the need to manually parse cryptic Kubernetes error messages by translating complex cluster states into actionable, human-readable diagnostics.

How K8sGPT Works

K8sGPT scans your Kubernetes cluster for common issues — pending pods, failing deployments, misconfigured services, resource quota violations, and more. It then sends the relevant context to an AI backend (OpenAI, Azure OpenAI, Google Gemini, Bedrock, LocalAI, or others) and returns a plain-language explanation with remediation suggestions.

Key Features

• 11+ AI backends: OpenAI, Azure OpenAI, Google Gemini, Amazon Bedrock, LocalAI, Ollama, and more
• Multi-resource analysis: Pods, deployments, services, ingresses, PVCs, and custom resources
• Kubernetes Operator: Deploy as an in-cluster operator for continuous scanning
• CLI tool: Simple k8sgpt analyze command for one-off diagnostics
• Custom analyzers: Extend with your own resource analysis logic
• Multi-language support: Explanations in multiple human languages
• CNCF Sandbox: Vendor-neutral governance under CNCF

Who Uses K8sGPT?

K8sGPT is used by Kubernetes operators, SREs, and DevOps engineers who want faster cluster troubleshooting. It's particularly valuable for teams with mixed Kubernetes experience levels, as it makes expert-level diagnostics accessible to everyone.

Pricing

K8sGPT is completely free and open source. AI backend costs depend on your chosen provider.