NVIDIA Vera Rubin NVL72 Leads Agentic Inference Performance on MLPerf Endpoints v1.0 NVIDIA reported provisional MLPerf Endpoints v1.0 results showing its Vera Rubin NVL72 delivers more than 10x the throughput per kilowatt of provisioned power of NVIDIA GB300 NVL72 at a shared minimum end-to-end interactivity of 50 tokens per second per user. In the Kimi K3 run published November 9, 2026, Vera Rubin NVL72 (32 accelerators) reached 41,414.2 system tokens/sec at 100.89 kW with 1,024 concurrent queries and a TTFT P90 of 4,673.9 ms, versus GB300 NVL72 (48 accelerators) at 2,210.4 tokens/sec and 60 kW with 48 queries. The Vera Rubin figures are pending MLCommons Association verification and peer review, while the GB300 results were verified by MLCommons. Agentic workloads are demanding more from inference systems. AI agents carry growing conversation histories across turns and generate a mix of reasoning, tool calls, and responses. Inference systems must support users at scale while remaining responsive. MLPerf Endpoints v1.0 evaluates this balance between serving capacity and responsiveness using coding and enterprise workflows through API endpoints. During performance runs, generated tool calls and workflow intent are checked against reference values. Separate SWE-bench Verified evaluations assess end-to-end task-solving capability. These checks evaluate whether faster serving preserves model quality. In MLPerf Endpoints v1.0, provisional NVIDIA Vera Rubin NVL72 results show over 10x the throughput per kilowatt of provisioned power of NVIDIA GB300 NVL72 at a shared minimum end-to-end interactivity of 50 tokens per second per user.