Google unveils open AI stack for Kubernetes

Google Cloud introduces an open-source AI-native stack for Kubernetes at KubeCon + CloudNativeCon North America 2026. The suite includes OpenRL, llm-d 1.0, Ray on GKE, Agent Substrate, gVisor, and TorchTPU to handle distributed inference and agents. llm-d 1.0 doubled KV cache hit rates from 35% to 70% and cut median time-to-first-token by 70-85% on multi-turn agent workloads.