MiniMax M3 428B — GB300 NVL72 vs H200 Performance per Dollar
Cost per million tokens of GB300 NVL72 (NVIDIA Blackwell) versus H200 (NVIDIA Hopper) on MiniMax M3 428B. 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.
Push MiniMax M3 428B to 39 tok/s/user and GB300 NVL72 lands at $0.37 per million tokens against H200's $0.30 — H200 pulls ahead by 23%.
GB300 NVL72: $1.50 per million tokens. H200: $0.54. Both at 73 tok/s/user on MiniMax M3 428B, with H200 180% cheaper.
Toward the upper edge of the 5–142 tok/s/user interactivity band — at 108 tok/s/user — GB300 NVL72 runs $3.70 per million tokens on MiniMax M3 428B while H200 runs $0.82. H200 is the cheaper choice by 352%. (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.)
GPU pricing (owning hyperscaler): GB300 NVL72 $2.65/GPU/hr · H200 $1.41/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 | GB300 NVL72:$0.371H200:$0.303 | GB300 NVL72:$1.505H200:$0.537 | GB300 NVL72:$3.696H200:$0.819 |
| Concurrency | GB300 NVL72:~287H200:~73 | GB300 NVL72:~24H200:~21 | GB300 NVL72:~5H200:~10 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.