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 101 tok/s/user and GB300 NVL72 lands at $0.06 per million tokens against H200's $0.29 — GB300 NVL72 pulls ahead by 389%.
GB300 NVL72: $0.29 per million tokens. H200: $0.41. Both at 180 tok/s/user on MiniMax M3 428B, with GB300 NVL72 41% cheaper.
Toward the upper edge of the 22–339 tok/s/user interactivity band — at 260 tok/s/user — GB300 NVL72 runs $0.38 per million tokens on MiniMax M3 428B while H200 runs $0.91. GB300 NVL72 is the cheaper choice by 139%. (Numbers reflect the default 8k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Chip pricing (owning hyperscaler): GB300 NVL72 $2.31/chip/hr · H200 $1.22/chip/hr. Source: SemiAnalysis Market July 2026 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:$0.059H200:$0.289 | GB300 NVL72:$0.291H200:$0.410 | GB300 NVL72:$0.380H200:$0.908 |
| Concurrency | GB300 NVL72:~257H200:~6 | GB300 NVL72:~12H200:~3 | GB300 NVL72:~8H200:~1 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.