Release Notes — 2.3.4

Released: 2026-05-10

Kure Monitor 2.3.4 refreshes the LLM-provider experience: the default provider for new installs switches from Ollama to Groq, and the model catalogs across all six providers are bumped to current (May 2026) latest models in both the frontend dropdowns and the backend default_model().

No RBAC or breaking changes. Drop-in upgrade. No operator action required. No schema migration. Existing provider configuration in the database is preserved untouched — only the defaults shown in the UI change.

Highlights

• Groq is the new default LLM provider. When a fresh install loads the Admin Panel → AI Configuration page, Groq is now pre-selected in the provider dropdown (previously Ollama). This only affects the initial pre-selection — once you save a configuration, your choice is sticky and no future release will silently change it.
• Model catalogs refreshed across all six providers. Both the frontend dropdown options and the backend default_model() for each provider were bumped to the May 2026 latest. Backend tests in backend/tests/test_llm_factory.py and backend/tests/test_copilot_provider.py assert the new model IDs.

Updated model lists

Defaults shown in bold.

Provider	2.3.4 models	Previous (2.3.3)
OpenAI	gpt-5.5, gpt-5.5-mini, gpt-5.4-mini	gpt-5, gpt-5-mini, gpt-4.1
Anthropic	claude-opus-4-7, claude-sonnet-4-6, claude-haiku-4-5	claude-opus-4-5, claude-sonnet-4-5, claude-haiku-4-5
Google Gemini	gemini-3.1-pro, gemini-3-flash, gemini-3.1-flash-lite	gemini-2.5-pro, gemini-2.5-flash, gemini-2.5-flash-lite
Groq	openai/gpt-oss-120b, meta-llama/llama-4-scout-17b-16e-instruct, llama-3.3-70b-versatile, openai/gpt-oss-20b	Maverick, Scout, llama-3.3-70b-versatile
Ollama	llama4:scout, llama3.3, qwen3	llama3.3, llama3.2, qwen2.5
GitHub Copilot (GitHub Models)	openai/gpt-5.5, openai/gpt-5.5-mini, anthropic/claude-sonnet-4-6	openai/gpt-5, openai/gpt-5-mini, anthropic/claude-sonnet-4

What’s changed

Changed

• Default LLM provider switched from Ollama to Groq in the AI Configuration panel for new installs (frontend/src/components/LLMSettings.js).
• Model lists refreshed across all six providers in frontend/src/components/LLMSettings.js (dropdown options) and backend/llm_providers/ (each provider’s default_model()).
• Backend unit tests in backend/tests/test_llm_factory.py and backend/tests/test_copilot_provider.py updated to assert the new model IDs.

Not changed

• No RBAC change. The backend ClusterRole is identical to 2.3.3.
• No schema migration. No database changes; existing LLM configuration rows are read back unchanged.
• No Helm values rename. All chart values keep their existing keys.
• Existing API keys preserved. Provider config stored in the database is untouched. If you already have a provider configured, this release changes nothing about which provider is active.

Upgrading

Helm

helm repo update
helm upgrade kure-monitor kure-monitor/kure \
  --namespace kure-system \
  --version 2.3.4

Raw k8s manifests

# Bump image tags in k8s/{backend,frontend,agent,security-scanner}.yaml to 2.3.4
kubectl apply -f k8s/

Verifying the upgrade

1. Open the dashboard.
2. Open Admin Panel → AI Configuration.
3. The dropdowns should now expose the refreshed model lists. If you have a provider already configured, your previously selected model is still in the dropdown (assuming it’s one of the new options) and remains the active model. If you previously had a model that has been retired from the catalog, pick a new one and click Test Connection → Save.

Full changelog

See CHANGELOG.md in the repository.