MiniMax M2.5/M2.7 — B300 vs GB300 NVL72
Head-to-head AI inference benchmark comparison of B300 (NVIDIA Blackwell) and GB300 NVL72 (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.
B300 posts 7135 tok/s/chip for $0.09 per million tokens at 69 tok/s/user on MiniMax M2.5/M2.7; GB300 NVL72 posts 8401 tok/s/chip for $0.07. GB300 NVL72 is 22% cheaper per token; GB300 NVL72 delivers 18% more tok/s/chip.
Throughput at 108 tok/s/user on MiniMax M2.5/M2.7: B300 hits 2787 tok/s/chip, GB300 NVL72 hits 3059. Per-million costs land at $0.23 and $0.21 respectively. GB300 NVL72 is 7% cheaper per token; GB300 NVL72 delivers 10% more tok/s/chip.
B300 / GB300 NVL72 on MiniMax M2.5/M2.7 at 147 tok/s/user: 1049 / 1009 tok/s/chip, $0.60 / $0.63 per million tokens. B300 is 6% cheaper per token; B300 delivers 4% more tok/s/chip. (Numbers reflect the default 1k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B300:7135.1GB300 NVL72:8401.0 | B300:2786.7GB300 NVL72:3058.6 | B300:1049.5GB300 NVL72:1008.6 |
| Cost ($/M tok) | B300:$0.089GB300 NVL72:$0.073 | B300:$0.225GB300 NVL72:$0.210 | B300:$0.599GB300 NVL72:$0.634 |
| tok/s/MW | B300:3755301GB300 NVL72:3962724 | B300:1466680GB300 NVL72:1442744 | B300:552360GB300 NVL72:475763 |
| Concurrency | B300:~581GB300 NVL72:~1024 | B300:~72GB300 NVL72:~121 | B300:~9GB300 NVL72:~22 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.