OOpenclawnet
  • Introduction
  • Docs
      • 20260425 concept review
      • 20260503 repo split decision
      • Agent runtime
      • Components
      • Glossary
      • Jobs
      • Memory service proposal
      • Nemoclaw mapping
      • Openclaw mapping
      • Overview
      • Provider model
      • Runtime flow
      • Scenarios s4 s5 plan
      • Secrets vault admin ui
      • Secrets vault azure readiness
      • Secrets vault evolution
      • Secrets vault lifecycle phase4
      • Secrets vault lifecycle phase5
      • Secrets vault phase5
      • Secrets vault phase6
      • Secrets vault threat model
      • Source of truth rules
      • Storage
  • Scripts
  • Sessions
    • RELEASE CHECKLIST
    • Session 2 guide es
    • Session 2 guide
    • Session 3 guide es
    • Session 3 guide
    • Session 4 guide es
    • Session 4 guide
    • Session 5 guide es
    • Session 5 guide
    • Speakers
  • Src
  • Tests
Powered by Docsbook
Docs/Architecture/Provider model
OverviewPreviousRuntime flowNext

On this page

  • Primary Interface — IAgentProvider
  • Available Providers
  • Multi-Instance Model Providers
  • Agent Profiles
  • Provider Resolution Flow
  • Provider Configuration
  • Ollama (Default — local, no API key)
  • FoundryLocal (local on-device inference)
  • Azure OpenAI (cloud)
  • Foundry (Azure AI cloud endpoint)
  • Per-Definition Endpoint Override
  • Example: Two Ollama Instances
  • How It Works
  • Fallback Chain
  • Configuration
  • How It Works
  • Availability Check
  • Recommended Chain Configurations
  • Model Selection Strategy
  • Primary Model
  • Per-Request Routing (Future)
  • Runtime Model Switching
  • Switching Providers
  • At Startup (DI Registration)
  • With Fallback Chain
  • At Runtime (Settings API)

Was this page helpful?