Head-to-head AI inference benchmark comparison of GB200 NVL72 (NVIDIA Blackwell) and H200 (NVIDIA Hopper) on MiniMax M3.1 Flash 428B. Latency, throughput, and cost across LLM workloads. Use the chart controls below to switch sequences, precisions, and metrics β same interactions as the main inference chart.
AgentX replays real coding-agent sessions rather than fixed-length prompts, so context grows turn over turn and most of each request is served from cache instead of being recomputed. That turns the comparison into a systems question: KV transfer between nodes, prefix-aware routing, and cache capacity all move the curve alongside raw chip throughput. Fixed-sequence workloads stay the clean baseline for kernel and silicon performance, so the two scenarios answer different questions about the same hardware. Learn more about AgentX β
GB200 NVL72 / H200 on MiniMax M3.1 Flash 428B at 81 tok/s/user: 52698 / 9854 tok/s/chip, $0.01 / $0.03 per million tokens. GB200 NVL72 is 251% cheaper per token; GB200 NVL72 delivers 435% more tok/s/chip.
Around the middle of the 10β294 tok/s/user interactivity band, at 152 tok/s/user on MiniMax M3.1 Flash 428B: GB200 NVL72 runs 43465 tok/s/chip at $0.01/M tokens, H200 runs 7133 at $0.05/M. GB200 NVL72 is 300% cheaper per token; GB200 NVL72 delivers 509% more tok/s/chip.
Setting 223 tok/s/user as the target on MiniMax M3.1 Flash 428B, GB200 NVL72 produces 28152 tok/s/chip ($0.02 per million tokens) and H200 produces 4231 ($0.08). GB200 NVL72 is 336% cheaper per token; GB200 NVL72 delivers 565% more tok/s/chip. (Numbers reflect the default agentic-traces Β· fp4 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB200 NVL72:52698.4H200:9854.4 | GB200 NVL72:43464.7H200:7132.8 | GB200 NVL72:28151.6H200:4231.4 |
| Cost ($/M tok) | GB200 NVL72:$0.010H200:$0.034 | GB200 NVL72:$0.012H200:$0.048 | GB200 NVL72:$0.018H200:$0.080 |
| tok/s/MW | GB200 NVL72:28180958H200:7192983 | GB200 NVL72:23243173H200:5206434 | GB200 NVL72:15054326H200:3088604 |
| Concurrency | GB200 NVL72:~32H200:~16 | GB200 NVL72:~26H200:~10 | GB200 NVL72:~16H200:~5 |
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
No measurements to plot for this selection. Review the benchmark controls above or adjust quick filters.
