Why AI agents Cost So Much

From the creator

AI agents can quickly deliver business value, but as usage scales, many organizations are discovering that intelligent experiences come with real operational costs. In this episode of the Azure Essentials Show, host Takuto Higuchi sits down with Sindhura Bandhakavi, Director of Engineering, Microsoft CoreAI, to explore the emerging discipline of AI token economics, or "tokenomics," and what it takes to optimize agent performance without sacrificing quality. Through practical demonstrations of capabilities such as Model Router and Agent Optimizer in Microsoft Foundry, they unpack the factors that drive AI costs, explain why reducing tokens alone is not always the right strategy, and show how organizations can treat AI as a managed investment. Whether you're building your first AI agent or looking to improve ROI across a growing portfolio of agents, this episode provides actionable guidance for balancing cost, performance, and scalability. Three things you will learn - Why AI agents can become expensive to operate at scale and the key factors that drive token consumption, model costs, and overall AI ROI. You'll learn why simply reducing tokens is not always the best optimization strategy. - How to use Azure AI Foundry capabilities such as Model Router and Agent Optimizer to improve agent efficiency, automatically select the right model for each request, and continuously enhance agent performance over time. - How to use Azure AI Foundry capabilities such as Model Route

Published 2026-10-20 · Watch on YouTube ↗
Availability and playback restrictions are controlled by the original channel.