Skip to main content

Feature Settings Configuration

The FeatureSetting model manages configurations for various functional scenarios, enabling flexible binding between scenarios and resources like LLMs.

Supported Scenarios​

ScenarioDescription
Knowledge GraphConfigure the LLM used for graph building and concept extraction
AI AssistantConfigure the LLM model used for assistant conversations
Content SummarizationConfigure the LLM used for automatic note summarization

Configuration Method​

  • Each scenario is configured independently with different LLM options
  • Managed via the /api/feature-settings API
  • The frontend provides a visual configuration interface in the settings page
  • Supports configuration grouping and dynamic expansion

Usage Example​

Suppose you have configured both GPT-4 and Claude:

  • Use GPT-4 for knowledge graph extraction (strong concept extraction capability)
  • Use Claude for AI assistant conversations (better conversational experience)
  • Use a lower-cost model for content summarization