Head-to-head AI inference benchmark comparison of B300 (NVIDIA Blackwell) and GB300 NVL72 (NVIDIA Blackwell) on GLM 5.3 744B. Latency, throughput, and cost across LLM workloads. Use the chart controls below to switch sequences, precisions, and metrics — 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 →
At 133 tok/s/user interactivity on GLM 5.3 744B, B300 delivers 12134 tok/s/chip at $0.05 per million tokens; GB300 NVL72 delivers 16870 tok/s/chip at $0.04. GB300 NVL72 is 36% cheaper per token; GB300 NVL72 delivers 39% more tok/s/chip at this point.
B300 posts 7378 tok/s/chip for $0.09 per million tokens at 180 tok/s/user on GLM 5.3 744B; GB300 NVL72 posts 12861 tok/s/chip for $0.05. GB300 NVL72 is 71% cheaper per token; GB300 NVL72 delivers 74% more tok/s/chip.
Throughput at 226 tok/s/user on GLM 5.3 744B: B300 hits 5308 tok/s/chip, GB300 NVL72 hits 9605. Per-million costs land at $0.12 and $0.07 respectively. GB300 NVL72 is 77% cheaper per token; GB300 NVL72 delivers 81% more tok/s/chip. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B300:12134.0GB300 NVL72:16870.5 | B300:7377.9GB300 NVL72:12861.5 | B300:5307.6GB300 NVL72:9605.1 |
| Cost ($/M tok) | B300:$0.052GB300 NVL72:$0.038 | B300:$0.085GB300 NVL72:$0.050 | B300:$0.118GB300 NVL72:$0.067 |
| tok/s/MW | B300:6386331GB300 NVL72:7957760 | B300:3883110GB300 NVL72:6066734 | B300:2793471GB300 NVL72:4530731 |
| Concurrency | B300:~15GB300 NVL72:~49 | B300:~10GB300 NVL72:~49 | B300:~6GB300 NVL72:~30 |
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.
