F5 gateway promises 60% lower token spend

F5 folded model routing, MCP governance, and prompt guardrails into its AI Gateway, claiming the combined control plane can reduce token spend by up to 60% without application changes. F5’s announcement also reports that 77% of organizations now call inference their dominant AI activity and manage an average of seven models. Platform teams should record the claim as a procurement benchmark, then demand workload-level evidence before adding another proxy to the stack—an update to the archive’s break-even test for AI gateways.