MiniMax M3 428B — B200 vs H200 Performance per Dollar
Cost per million tokens of B200 (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.
On MiniMax M3 428B at 56 tok/s/user, the per-million math comes out to $0.57 for B200 and $0.43 for H200; H200 delivers 31% more output per dollar.
At 103 tok/s/user on MiniMax M3 428B, B200 costs $1.57 per million tokens; H200 costs $0.77. H200 is 105% more cost-efficient at this operating point.
H200 edges B200 at 151 tok/s/user on MiniMax M3 428B — $1.15 per million tokens versus $4.32, a 274% cost-per-token gap. (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): B200 $1.95/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 | B200:$0.567H200:$0.433 | B200:$1.572H200:$0.767 | B200:$4.315H200:$1.153 |
| Concurrency | B200:~40H200:~35 | B200:~8H200:~11 | B200:~4H200:~4 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.