Route LLM requests to cheapest operational model
Automatically dispatches AI requests to the most cost-effective and available provider, reducing inference costs while maintaining service reliability through real-time status checks.
How it works
The agent calls the cheapest_model tool with specific context window and operational constraints. It analyzes the ranked list of providers to identify the top option. Before dispatching, it verifies the provider's operational status. The agent then routes the request to the selected provider's API endpoint. This process can be repeated periodically or triggered by price change alerts to ensure continuous cost optimization.
Prompt
Paste this into an agent that has the servers below connected. Adjust the names and dates to your own.
Find the cheapest operational LLM model that supports at least 128K context window. Check its current status. If it is operational, provide the provider name and the estimated cost per million tokens. Then, simulate routing a request to this provider.MCP servers it needs
Source
The vendor page this playbook is drawn from, and the passage that supports it.
Programmatically route to the cheapest operational LLM, read exclusive per-model price history, check provider status, and subscribe to price-cut alertsAI API Radar — Developer & Agent API