MiniMax M2.5/M2.7 · GPU comparison

MiniMax M2.5/M2.7 — GB200 NVL72 vs H200

Head-to-head AI inference benchmark comparison of GB200 NVL72 (NVIDIA Blackwell) and H200 (NVIDIA Hopper) on MiniMax M2.5/M2.7. Latency, throughput, and cost across LLM workloads. Use the chart controls below to switch sequences, precisions, and metrics — same interactions as the main inference chart.

GB200 NVL72 posts 3623 tok/s/GPU for $0.17 per million tokens at 52 tok/s/user on MiniMax M2.5/M2.7; H200 posts 1177 tok/s/GPU for $0.34. GB200 NVL72 is 99% cheaper per token; GB200 NVL72 delivers 208% more tok/s/GPU.

Throughput at 80 tok/s/user on MiniMax M2.5/M2.7: GB200 NVL72 hits 1276 tok/s/GPU, H200 hits 572. Per-million costs land at $0.50 and $0.68 respectively. GB200 NVL72 is 37% cheaper per token; GB200 NVL72 delivers 123% more tok/s/GPU.

GB200 NVL72 / H200 on MiniMax M2.5/M2.7 at 107 tok/s/user: 667 / 307 tok/s/GPU, $0.92 / $1.27 per million tokens. GB200 NVL72 is 37% cheaper per token; GB200 NVL72 delivers 117% more tok/s/GPU. (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.)

View performance-per-dollar view →

Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.
Metric
Interactivity (tok/s/user)
Interactivity (tok/s/user)
Interactivity (tok/s/user)
Throughput (tok/s/gpu)
GB200 NVL72:3622.7H200:1176.5
GB200 NVL72:1276.4H200:572.2
GB200 NVL72:667.3H200:307.0
Cost ($/M tok)
GB200 NVL72:$0.169H200:$0.337
GB200 NVL72:$0.497H200:$0.683
GB200 NVL72:$0.925H200:$1.271
tok/s/MW
GB200 NVL72:1725082H200:680077
GB200 NVL72:607826H200:330764
GB200 NVL72:317769H200:177472
Concurrency
GB200 NVL72:~489H200:~106
GB200 NVL72:~64H200:~14
GB200 NVL72:~32H200:~5

Inference Performance

Inference performance metrics across different models, hardware configurations, and serving parameters.