MiniMax M2.5/M2.7 — B300 vs H100
Head-to-head AI inference benchmark comparison of B300 (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.
B300 / H100 on MiniMax M2.5/M2.7 at 59 tok/s/user: 3396 / 628 tok/s/GPU, $0.19 / $0.57 per million tokens. B300 is 199% cheaper per token; B300 delivers 441% more tok/s/GPU.
Around the middle of the 40–116 tok/s/user interactivity band, at 78 tok/s/user on MiniMax M2.5/M2.7: B300 runs 1799 tok/s/GPU at $0.37/M tokens, H100 runs 376 at $0.94/M. B300 is 156% cheaper per token; B300 delivers 379% more tok/s/GPU.
Setting 97 tok/s/user as the target on MiniMax M2.5/M2.7, B300 produces 1050 tok/s/GPU ($0.59 per million tokens) and H100 produces 208 ($1.75). B300 is 198% cheaper per token; B300 delivers 405% more tok/s/GPU. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | B300:3395.9H100:628.0 | B300:1799.0H100:375.6 | B300:1050.1H100:207.9 |
| Cost ($/M tok) | B300:$0.191H100:$0.572 | B300:$0.367H100:$0.939 | B300:$0.588H100:$1.754 |
| tok/s/MW | B300:1564951H100:363025 | B300:829013H100:217088 | B300:483894H100:120199 |
| Concurrency | B300:~118H100:~43 | B300:~47H100:~12 | B300:~22H100:~8 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.