Microsoft Foundry Slow? Check These 12 Response Headers

Slow requests to Microsoft Foundry? Response headers can help you investigate what's happening behind the scenes. In this video, I walk through 12 useful response headers and show how to capture them with the OpenAI libraries and Microsoft Agent Framework, using both streaming and non-streaming requests. You'll learn how to inspect remaining request and token limits, processing time, region and backend cluster information, and request identifiers that can help when investigating an issue or opening a support ticket. I also show the Foundry tracing view and discuss a performance difference I observed between the Responses API and Chat Completions API. Chapters 00:00 Investigating slow Foundry requests 00:39 12 response headers for troubleshooting 01:28 Microsoft Agent Framework and OpenAI examples 02:25 Running streaming and non-streaming requests 02:44 Reading limits, timing and backend details 04:08 Looking at traces in Foundry 04:59 Responses API vs. Chat Completions observations 05:22 Wrap-up Which Foundry performance issues have you run into? Let me know in the comments. #MicrosoftFoundry #AgentFramework #OpenAI #Troubleshooting