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 44 tok/s/user: 9991 / 1891 tok/s/chip, $0.06 / $0.17 per million tokens. B300 is 173% cheaper per token; B300 delivers 428% more tok/s/chip.
Around the middle of the 21–111 tok/s/user interactivity band, at 66 tok/s/user on MiniMax M2.5/M2.7: B300 runs 5508 tok/s/chip at $0.11/M tokens, H100 runs 1269 at $0.26/M. B300 is 125% cheaper per token; B300 delivers 334% more tok/s/chip.
Setting 89 tok/s/user as the target on MiniMax M2.5/M2.7, B300 produces 4006 tok/s/chip ($0.16 per million tokens) and H100 produces 806 ($0.40). B300 is 157% cheaper per token; B300 delivers 397% more tok/s/chip. (Numbers reflect the default 8k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B300:9990.5H100:1891.3 | B300:5508.0H100:1268.8 | B300:4006.4H100:805.9 |
| Cost ($/M tok) | B300:$0.063H100:$0.172 | B300:$0.114H100:$0.256 | B300:$0.157H100:$0.403 |
| tok/s/MW | B300:5258170H100:1380510 | B300:2898963H100:926145 | B300:2108613H100:588262 |
| Concurrency | B300:~190H100:~39 | B300:~30H100:~18 | B300:~4H100:~9 |
Inference Performance
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
Total Tokens per $1 USD (Owning - Hyperscaler) vs. Interactivity
MiniMax M2.5/2.7 230B • FP4 • 8K / 1K • Source: SemiAnalysis InferenceX™
TCO $/chip/hr: H100: 1.17H200: 1.22B200: 1.73B300: 2.26GB200: 1.86GB300: 2.31MI300X: 0.95MI325X: 1.1MI355X: 1.5RTX6000PRO: 0.68
Source: SemiAnalysis Market July 2026 Pricing Surveys & AI Cloud TCO Model
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate cost per million tokens per decode chip or per prefill chip, rather than per total chip count. This makes direct token cost comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate input throughput per decode chip or per prefill chip, rather than per total chip count. This makes direct input throughput comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate output throughput per decode chip or per prefill chip, rather than per total chip count. This makes direct output throughput comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate power per decode chip or per prefill chip, rather than per total chip count. This makes direct power comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate Joules per decode chip or per prefill chip, rather than per total chip count. This makes direct Joules per token comparison with aggregated configs not an apples-to-apples comparison.
No data available
Please change the model, sequence, precision, date range or chip selection.
Shift+Scroll to zoom • Drag to pan • Double-click to reset • Click a point to pin tooltip