MiniMax M3 428B — B200 vs MI300X Performance per Dollar
Cost per million tokens of B200 (NVIDIA Blackwell) versus MI300X (AMD CDNA 3) 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.
Near the low end of the 15–149 tok/s/user interactivity band — at 48 tok/s/user — B200 runs $0.15 per million tokens on MiniMax M3 428B while MI300X runs $0.51. B200 is the cheaper choice by 245%.
On MiniMax M3 428B at 82 tok/s/user, the per-million math comes out to $0.29 for B200 and $1.31 for MI300X; B200 delivers 354% more output per dollar.
At 115 tok/s/user on MiniMax M3 428B, B200 costs $0.36 per million tokens; MI300X costs $1.82. B200 is 401% more cost-efficient at this operating point. (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.)
Chip pricing (owning hyperscaler): B200 $1.73/chip/hr · MI300X $0.95/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 | B200:$0.148MI300X:$0.509 | B200:$0.288MI300X:$1.307 | B200:$0.363MI300X:$1.819 |
| Concurrency | B200:~168MI300X:~45 | B200:~62MI300X:~10 | B200:~28MI300X:~5 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.