GLM 5/5.1 — B200 vs GB300 NVL72
Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and GB300 NVL72 (NVIDIA Blackwell) on GLM 5/5.1. 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.
Throughput at 31 tok/s/user on GLM 5/5.1: B200 hits 669 tok/s/GPU, GB300 NVL72 hits 3824. Per-million costs land at $0.82 and $0.19 respectively. GB300 NVL72 is 325% cheaper per token; GB300 NVL72 delivers 471% more tok/s/GPU.
B200 / GB300 NVL72 on GLM 5/5.1 at 43 tok/s/user: 444 / 276 tok/s/GPU, $1.23 / $2.67 per million tokens. B200 is 118% cheaper per token; B200 delivers 61% more tok/s/GPU.
Toward the upper edge of the 19–67 tok/s/user interactivity band, at 55 tok/s/user on GLM 5/5.1: B200 runs 382 tok/s/GPU at $1.42/M tokens, GB300 NVL72 runs 93 at $7.90/M. B200 is 457% cheaper per token; B200 delivers 310% more tok/s/GPU. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | B200:669.2GB300 NVL72:3824.1 | B200:443.9GB300 NVL72:275.7 | B200:381.9GB300 NVL72:93.2 |
| Cost ($/M tok) | B200:$0.820GB300 NVL72:$0.193 | B200:$1.225GB300 NVL72:$2.674 | B200:$1.420GB300 NVL72:$7.904 |
| tok/s/MW | B200:308378GB300 NVL72:1820984 | B200:204554GB300 NVL72:131293 | B200:176008GB300 NVL72:44379 |
| Concurrency | B200:~450GB300 NVL72:~6137 | B200:~393GB300 NVL72:~258 | B200:~28GB300 NVL72:~67 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.