Am I paying for a giant model I don't need?
Microsoft Foundry model router achieves 37.7% cost savings by routing prompts to appropriately-sized models
Microsoft's Azure AI Foundry model router automatically routes agent prompts to the most cost-appropriate model within a family (e.g., GPT 5.6 Luna/Terra/Soul), with an open-source auto-evaluation tool to benchmark quality, cost, and latency tradeoffs before committing. A demo showed 37.7% cost savings vs. using the top-tier model exclusively. This matters as enterprises increasingly look to optimize inference spend without sacrificing agent quality.