AI Models
Configure LLM providers — keys, models, base URLs, and the active provider.
The AI Models tab (nevoflux://settings/llm) is where you connect LLM
providers. This is the tab the first-run setup sends you to.
Layout
Providers are shown as cards in two groups:
- LLM Providers — cloud APIs (Anthropic, OpenAI, DeepSeek, Qwen, Gemini, Groq, OpenRouter, Mistral, xAI/Grok, Cohere, Perplexity, Together) and local (Ollama).
- Agents — command-line/agent providers (Claude Code, Gemini CLI, Kimi Agent, OpenClaw).
A card shows an Active badge for the current provider and a checkmark when it is configured.
Configure a provider
Click a card to open the Configure Provider dialog:
| Field | Notes |
|---|---|
| API Key | Stored in config.toml; shows "configured" once set |
| Model | Optional; the provider's default is shown as a hint |
| Base URL | Optional; for proxies / compatible endpoints (hidden for CLI providers) |
| Set as active provider | Make this the provider the agent uses |
Saving writes [llm.<provider>] (api_key, model, base_url) into
config.toml and, if ticked, sets it active.
Example: click Anthropic, paste your key, optionally set
claude-sonnet-4-20250514 as the model, tick Set as active provider, Save.
The sidebar can now answer.
See Configuration → Providers & API keys for the full provider table and the environment variables you can use instead of storing keys here.