Platform

Intelligent Model Arbiter

Every LLM request is scored in real time against all models in your tier. The Arbiter selects the cheapest model that meets your quality and latency SLA — automatically, on every call.

How it works

The Arbiter runs before your request reaches any provider. It evaluates cost, output quality (measured against your golden dataset), and p99 latency for each eligible model, then routes to the optimal one.

Configuration

Set a routing hint per call or configure a default globally in Dashboard → Arbitration.

const response = await openai.chat.completions.create({
  model: 'gpt-4o',          // preferred model — Arbiter may route elsewhere
  messages,
  azoth_routing: 'cost',    // 'cost' | 'quality' | 'latency' | 'balanced'
});

Routing modes