InferenceX
中文
MiniMax M2.5/M2.7 · Chip comparison

MiniMax M2.5/M2.7 — GB200 NVL72 vs H100

Head-to-head AI inference benchmark comparison of GB200 NVL72 (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.

At 50 tok/s/user interactivity on MiniMax M2.5/M2.7, GB200 NVL72 delivers 8615 tok/s/chip at $0.06 per million tokens; H100 delivers 1692 tok/s/chip at $0.19. GB200 NVL72 is 220% cheaper per token; GB200 NVL72 delivers 409% more tok/s/chip at this point.

GB200 NVL72 posts 4540 tok/s/chip for $0.11 per million tokens at 71 tok/s/user on MiniMax M2.5/M2.7; H100 posts 1155 tok/s/chip for $0.28. GB200 NVL72 is 147% cheaper per token; GB200 NVL72 delivers 293% more tok/s/chip.

Throughput at 91 tok/s/user on MiniMax M2.5/M2.7: GB200 NVL72 hits 2725 tok/s/chip, H100 hits 772. Per-million costs land at $0.19 and $0.42 respectively. GB200 NVL72 is 122% cheaper per token; GB200 NVL72 delivers 253% 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 view →

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)
GB200 NVL72:8615.0H100:1692.4
GB200 NVL72:4540.3H100:1154.7
GB200 NVL72:2725.1H100:771.8
Cost ($/M tok)
GB200 NVL72:$0.060H100:$0.192
GB200 NVL72:$0.114H100:$0.281
GB200 NVL72:$0.190H100:$0.421
tok/s/MW
GB200 NVL72:4606957H100:1235349
GB200 NVL72:2427969H100:842881
GB200 NVL72:1457292H100:563337
Concurrency
GB200 NVL72:~130H100:~31
GB200 NVL72:~47H100:~15
GB200 NVL72:~22H100:~8

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 - Neocloud Giant) vs. Interactivity

MiniMax M2.5/2.7 230B • FP4 • 8K / 1K • Source: SemiAnalysis InferenceX™

TCO $/chip/hr: VR200: 3.61H100: 1.55H200: 1.59B200: 2.07B300: 2.52GB200: 2.26GB300: 2.79MI300X: 1.16MI325X: 1.32MI355X: 2.09RTX6000PRO: 0.75Jalapeño (Teacup): 1.56

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