IBM (IBM) says LLM-D can reduce costs by as much as 5-10x and increase throughput by up to 3x for agent workloads requiring large context

IBM’s reported LLM-D claims include conditional cost and throughput gains; a separate announcement describes a benchmark demonstration and reports its workload-specific results.