WWDC26: Debug and profile agentic app experiences with Instruments | Apple
Apple's Instruments now profiles Foundation Models apps, giving developers observability into on-device and server LLM agent loops.
“Traditional code is predictable. LLMs are non-deterministic - the same input can produce different outputs.”
At WWDC26, Apple introduced a Foundation Models instrument in Xcode's Instruments to debug and profile agentic app experiences, exposing session structure, model inference latencies, tool calls, and token usage. It reflects Apple's push to make non-deterministic, multi-model LLM pipelines observable and debuggable for third-party developers, but is a tooling-focused developer session rather than a major industry-shaping announcement.