System diagnostics and provider routing for the OS reasoning engine.
| Model | Status | Latency | Route | Usage |
|---|---|---|---|---|
| Gemini 1.5 Pro | Active | 12ms | Core Reasoning | 120 / 500 |
| OpenAI gpt-4o | Fallback | 85ms | Complex Abstraction | 45 / 100 |
| Deepgram Nova-2 | Active | 25ms | Voice & Audio | 3 / 50 |
| Llama 3 70B | Offline | - | Local Fast-Path | 0 / 0 |
All API keys are encrypted at rest using AES-256-GCM and never logged in our systems.