MiniMax M2.5/M2.7 · GPU comparison

MiniMax M2.5/M2.7 — B200 vs H100

Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and H100 (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.

At 44 tok/s/user interactivity on MiniMax M2.5/M2.7, B200 delivers 9526 tok/s/GPU at $0.06 per million tokens; H100 delivers 1869 tok/s/GPU at $0.19. B200 is 238% cheaper per token; B200 delivers 410% more tok/s/GPU at this point.

B200 posts 4753 tok/s/GPU for $0.11 per million tokens at 66 tok/s/user on MiniMax M2.5/M2.7; H100 posts 1251 tok/s/GPU for $0.29. B200 is 155% cheaper per token; B200 delivers 280% more tok/s/GPU.

Throughput at 89 tok/s/user on MiniMax M2.5/M2.7: B200 hits 3180 tok/s/GPU, H100 hits 831. Per-million costs land at $0.17 and $0.44 respectively. B200 is 158% cheaper per token; B200 delivers 283% more tok/s/GPU. (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.)

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)
B200:9525.9H100:1868.7
B200:4752.7H100:1251.2
B200:3179.8H100:830.9
Cost ($/M tok)
B200:$0.057H100:$0.192
B200:$0.114H100:$0.291
B200:$0.170H100:$0.439
tok/s/MW
B200:4389834H100:1080187
B200:2190189H100:723258
B200:1465336H100:480286
Concurrency
B200:~512H100:~19
B200:~21H100:~16
B200:~20H100:~4

Inference Performance

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