MiniMax M2.5/M2.7 — B200 vs B300
Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and B300 (NVIDIA Blackwell) 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 68 tok/s/user on MiniMax M2.5/M2.7: B200 hits 6154 tok/s/GPU, B300 hits 6232. Per-million costs land at $0.09 and $0.10 respectively. B200 is 18% cheaper per token; B300 delivers 1% more tok/s/GPU.
B200 / B300 on MiniMax M2.5/M2.7 at 111 tok/s/user: 2085 / 2082 tok/s/GPU, $0.26 / $0.32 per million tokens. B200 is 20% cheaper per token; throughput per GPU is essentially tied.
Toward the upper edge of the 26–196 tok/s/user interactivity band, at 154 tok/s/user on MiniMax M2.5/M2.7: B200 runs 852 tok/s/GPU at $0.64/M tokens, B300 runs 895 at $0.74/M. B200 is 14% cheaper per token; B300 delivers 5% more tok/s/GPU. (Numbers reflect the default 1k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | B200:6154.4B300:6231.9 | B200:2084.9B300:2082.1 | B200:852.1B300:894.7 |
| Cost ($/M tok) | B200:$0.087B300:$0.103 | B200:$0.264B300:$0.317 | B200:$0.644B300:$0.736 |
| tok/s/MW | B200:2836134B300:2871858 | B200:960775B300:959506 | B200:392653B300:412310 |
| Concurrency | B200:~128B300:~128 | B200:~16B300:~17 | B200:~13B300:~6 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.