MiniMax M2.5/M2.7 · Chip comparison

MiniMax M2.5/M2.7 — B300 vs H200

Head-to-head AI inference benchmark comparison of B300 (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.

Throughput at 43 tok/s/user on MiniMax M2.5/M2.7: B300 hits 10,382 tok/s/chip, H200 hits 2,816. Costs per million tokens land at $0.060 and $0.12 respectively. B300 is 50% cheaper per token; B300 delivers 3.7× the tok/s/chip.

B300 / H200 on MiniMax M2.5/M2.7 at 72 tok/s/user: 4,787 / 1,855 tok/s/chip, $0.13 / $0.18 per million tokens. B300 is 28% cheaper per token; B300 delivers 158% more tok/s/chip.

Toward the upper edge of the 14–131 tok/s/user interactivity band, at 102 tok/s/user on MiniMax M2.5/M2.7: B300 runs 2,425 tok/s/chip at $0.26 per million tokens, H200 runs 1,011 at $0.34. B300 is 23% cheaper per token; B300 delivers 140% more tok/s/chip. (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 →

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/chip)
B30010,382H2002,816
B3004,787H2001,855
B3002,425H2001,011
Cost ($/M tok)
B300$0.060H200$0.12
B300$0.13H200$0.18
B300$0.26H200$0.34
tok/s/MW
B3005,464,464H2002,055,797
B3002,519,239H2001,354,090
B3001,276,295H200738,156
Concurrency
B300~215H200~30
B300~8H200~12
B300~16H200~5