Head-to-head AI inference benchmark comparison of GB300 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
Near the low end of the 112β335 tok/s/user interactivity band, at 168 tok/s/user on GLM 5.3 744B: GB300 NVL72 runs 14110 tok/s/chip at $0.05/M tokens, Vera Rubin NVL72 runs 77097 at $0.01/M. Vera Rubin NVL72 is 250% cheaper per token; Vera Rubin NVL72 delivers 446% more tok/s/chip.
Setting 224 tok/s/user as the target on GLM 5.3 744B, GB300 NVL72 produces 9956 tok/s/chip ($0.06 per million tokens) and Vera Rubin NVL72 produces 44770 ($0.02). Vera Rubin NVL72 is 188% cheaper per token; Vera Rubin NVL72 delivers 350% more tok/s/chip.
At 280 tok/s/user interactivity on GLM 5.3 744B, GB300 NVL72 delivers 3418 tok/s/chip at $0.19 per million tokens; Vera Rubin NVL72 delivers 26055 tok/s/chip at $0.04. Vera Rubin NVL72 is 388% cheaper per token; Vera Rubin NVL72 delivers 662% more tok/s/chip at this point. (Numbers reflect the default agentic-traces Β· fp4 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB300 NVL72:14110.3Vera Rubin NVL72:77097.4 | GB300 NVL72:9956.2Vera Rubin NVL72:44770.1 | GB300 NVL72:3417.5Vera Rubin NVL72:26055.2 |
| Cost ($/M tok) | GB300 NVL72:$0.045Vera Rubin NVL72:$0.013 | GB300 NVL72:$0.064Vera Rubin NVL72:$0.022 | GB300 NVL72:$0.188Vera Rubin NVL72:$0.038 |
| tok/s/MW | GB300 NVL72:6655779Vera Rubin NVL72:23362837 | GB300 NVL72:4696315Vera Rubin NVL72:13566704 | GB300 NVL72:1612051Vera Rubin NVL72:7895522 |
| Concurrency | GB300 NVL72:~60Vera Rubin NVL72:~265 | GB300 NVL72:~28Vera Rubin NVL72:~122 | GB300 NVL72:~3Vera Rubin NVL72:~50 |
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.
