Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and GB200 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.
Setting 67 tok/s/user as the target on MiniMax M2.5/M2.7, B200 produces 12576 tok/s/chip ($0.04 per million tokens) and GB200 NVL72 produces 14306 ($0.04). GB200 NVL72 is 6% cheaper per token; GB200 NVL72 delivers 14% more tok/s/chip.
At 100 tok/s/user interactivity on MiniMax M2.5/M2.7, B200 delivers 6803 tok/s/chip at $0.07 per million tokens; GB200 NVL72 delivers 7909 tok/s/chip at $0.07. GB200 NVL72 is 8% cheaper per token; GB200 NVL72 delivers 16% more tok/s/chip at this point.
B200 posts 4028 tok/s/chip for $0.12 per million tokens at 133 tok/s/user on MiniMax M2.5/M2.7; GB200 NVL72 posts 4076 tok/s/chip for $0.13. B200 is 6% cheaper per token; GB200 NVL72 delivers 1% more tok/s/chip. (Numbers reflect the default 8k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B200:12576.5GB200 NVL72:14306.4 | B200:6803.3GB200 NVL72:7909.2 | B200:4027.5GB200 NVL72:4075.6 |
| Cost ($/M tok) | B200:$0.038GB200 NVL72:$0.036 | B200:$0.071GB200 NVL72:$0.065 | B200:$0.119GB200 NVL72:$0.127 |
| tok/s/MW | B200:7354659GB200 NVL72:7650498 | B200:3978550GB200 NVL72:4229523 | B200:2355276GB200 NVL72:2179488 |
| Concurrency | B200:~1015GB200 NVL72:~909 | B200:~8GB200 NVL72:~46 | B200:~10GB200 NVL72:~19 |
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
TCO $/chip/hr: VR200: 3.61H100: 1.17H200: 1.22B200: 1.73B300: 2.26GB200: 1.86GB300: 2.31MI300X: 0.95MI325X: 1.1MI355X: 1.5RTX6000PRO: 0.68Jalapeño (Teacup): 1.47
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.
Matching measurements exist, but their chip series are hidden.
Shift+Scroll to zoom • Drag to pan • Double-click to reset • Click a point to pin tooltip
