Release Notes — 2.4.3
Released: 2026-06-15
Kure Monitor 2.4.3 focuses on unleashing the power of local and custom AI models, paired with significant UI and branding polish.
Headline Changes
Section titled “Headline Changes”- Local AI Auto-Discovery: Kure Monitor now automatically scans your Kubernetes cluster’s services for local LLMs (like Ollama, vLLM, LocalAI, TGI, or NVIDIA NIM). With a single click in the Admin panel, it automatically retrieves the service URL and lists the available models you have loaded.
- Flexible Custom AI Provider: We completely revamped the AI configuration. Rigid dropdowns for Ollama, OpenAI, and Groq have been merged into a unified, powerful “Custom” provider. You can now use any OpenAI-compatible endpoint with optional API keys, and type in absolutely any arbitrary model name you want (e.g.,
mistral:latest,llama-3.3-70b-versatile). - UI & Branding Updates: The official Kure Monitor logo has been integrated into the dashboard header, login screens, and documentation. To ensure a cohesive design, we’ve strictly enforced the Light Theme, removing the faulty dark theme toggle.
- NodeJS 24 & Helm Pipelines: Validated the frontend and backend against NodeJS 24 environments and updated
helmreleaseconfigurations to reflect the latest repository structure. - Network Policy Simplification: Completely removed all restrictive NetworkPolicies from the Helm chart. This provides immediate, frictionless internal network access for the Kure agent and backend, preventing timeouts when connecting to locally hosted AI models.
Upgrade Instructions
Section titled “Upgrade Instructions”As always, upgrading is a drop-in replacement via Helm:
helm repo updatehelm upgrade kure-monitor kure-monitor/kure \ --namespace kure-system \ --version 2.4.3No values changes are required. The new models are available in the LLM Settings tab for administrators.
Verifying the upgrade
Section titled “Verifying the upgrade”- Open the UI, you should immediately see the new Kure Monitor logo in the header instead of the generic server icon.
- Go to Admin → LLM Settings.
- Cycle through the providers (OpenAI, Anthropic, Gemini, Groq, Ollama, Copilot) and confirm the latest 2026 models are available in the dropdowns.
Full changelog
Section titled “Full changelog”See CHANGELOG.md in the repository.