Head-to-head AI inference benchmark comparison of GB200 NVL72 (NVIDIA Blackwell) and Vera Rubin NVL72 (NVIDIA Vera Rubin) on GLM 5.3 744B. Latency, throughput, and cost across LLM workloads. Use the chart controls below to switch sequences, precisions, and metrics β same interactions as the main inference chart.
AgentX replays real coding-agent sessions rather than fixed-length prompts, so context grows turn over turn and most of each request is served from cache instead of being recomputed. That turns the comparison into a systems question: KV transfer between nodes, prefix-aware routing, and cache capacity all move the curve alongside raw chip throughput. Fixed-sequence workloads stay the clean baseline for kernel and silicon performance, so the two scenarios answer different questions about the same hardware. Learn more about AgentX β
Setting 160 tok/s/user as the target on GLM 5.3 744B, GB200 NVL72 produces 12343 tok/s/chip ($0.04 per million tokens) and Vera Rubin NVL72 produces 82427 ($0.01). Vera Rubin NVL72 is 244% cheaper per token; Vera Rubin NVL72 delivers 568% more tok/s/chip.
At 208 tok/s/user interactivity on GLM 5.3 744B, GB200 NVL72 delivers 6947 tok/s/chip at $0.07 per million tokens; Vera Rubin NVL72 delivers 52921 tok/s/chip at $0.02. Vera Rubin NVL72 is 292% cheaper per token; Vera Rubin NVL72 delivers 662% more tok/s/chip at this point.
GB200 NVL72 posts 4397 tok/s/chip for $0.12 per million tokens at 256 tok/s/user on GLM 5.3 744B; Vera Rubin NVL72 posts 31935 tok/s/chip for $0.03. Vera Rubin NVL72 is 274% cheaper per token; Vera Rubin NVL72 delivers 626% more tok/s/chip. (Numbers reflect the default agentic-traces Β· fp4 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB200 NVL72:12343.0Vera Rubin NVL72:82426.8 | GB200 NVL72:6947.0Vera Rubin NVL72:52921.3 | GB200 NVL72:4397.1Vera Rubin NVL72:31935.5 |
| Cost ($/M tok) | GB200 NVL72:$0.042Vera Rubin NVL72:$0.012 | GB200 NVL72:$0.074Vera Rubin NVL72:$0.019 | GB200 NVL72:$0.118Vera Rubin NVL72:$0.031 |
| tok/s/MW | GB200 NVL72:6600558Vera Rubin NVL72:24977815 | GB200 NVL72:3715000Vera Rubin NVL72:16036744 | GB200 NVL72:2351386Vera Rubin NVL72:9677421 |
| Concurrency | GB200 NVL72:~56Vera Rubin NVL72:~290 | GB200 NVL72:~8Vera Rubin NVL72:~157 | GB200 NVL72:~4Vera Rubin NVL72:~71 |
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
No measurements to plot for this selection. Review the benchmark controls above or adjust quick filters.
