GLM 5/5.1 — B300 vs GB200 NVL72 Performance per Dollar
Cost per million tokens of B300 (NVIDIA Blackwell) versus GB200 NVL72 (NVIDIA Blackwell) on GLM 5/5.1. Owning-hyperscaler TCO normalized by output tokens — performance per dollar across LLM workloads. Pick the more cost-efficient SKU at every target interactivity level. Use the chart controls below to switch sequences, precisions, and metrics — same interactions as the main inference chart.
B300: $0.28 per million tokens. GB200 NVL72: $0.09. Both at 66 tok/s/user on GLM 5/5.1, with GB200 NVL72 196% cheaper.
Around the middle of the 40–147 tok/s/user interactivity band — at 93 tok/s/user — B300 runs $0.39 per million tokens on GLM 5/5.1 while GB200 NVL72 runs $0.31. GB200 NVL72 is the cheaper choice by 27%.
On GLM 5/5.1 at 120 tok/s/user, the per-million math comes out to $0.52 for B300 and $0.54 for GB200 NVL72; B300 delivers 4% more output per dollar. (Numbers reflect the default 8k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
GPU pricing (owning hyperscaler): B300 $2.34/GPU/hr · GB200 NVL72 $2.21/GPU/hr. Source: SemiAnalysis Market August 2025 Pricing Surveys & AI Cloud TCO Model.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Dollar per Million Tokens | B300:$0.280GB200 NVL72:$0.094 | B300:$0.393GB200 NVL72:$0.309 | B300:$0.518GB200 NVL72:$0.540 |
| Concurrency | B300:~17GB200 NVL72:~838 | B300:~9GB200 NVL72:~82 | B300:~4GB200 NVL72:~67 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.