Nexus One AI 🔔 Basic Tier
Intelligence Infrastructure

Model Router

Define routing rules that automatically send requests to the right model — fast models for simple queries, powerful models for reasoning, specialist models for vision and embeddings.

📡 Routes Active
of configured
🚀 Requests Routed
last 24 hours
⚡ Avg Latency Saved
vs. always using large model
☁️ Cloud Fallback
🔀
How Routing Works
💬
Request
User query
🔀
Router
Match rules
🖼️
Vision
Images/docs
·
🧠
Reasoning
Complex tasks
·
Fast
Simple Q&A
·
📚
Embedding
RAG / search
·
☁️
Cloud
Fallback only

Rules are evaluated in priority order (top → bottom). The first matching rule wins. Requests that match no rule go to the Default model.

📋
Routing Rules
Drag to reorder priority
🔩
Fallback & Cloud Settings
🏠
Default (Local) Model
Used when no routing rule matches. Always stays on-premises.
☁️
Cloud Fallback
Route to a cloud provider when local models are overloaded or unavailable. Disabled by default — enable only if data privacy policy permits.
🔒
PII Shield
Detect and redact personally identifiable information before routing to any cloud provider.
🧪
Live Routing Test
See which route would be selected for a given input
Routing Decision
Enter a prompt and click Test Routing to see which rule matches.