Head-to-head AI inference benchmark comparison of GB200 NVL72 (NVIDIA Blackwell) and H100 (NVIDIA Hopper) on DeepSeek V4.1 Flash 552B. 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.
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 →
GB200 NVL72 posts 15745 tok/s/chip for $0.03 per million tokens at 226 tok/s/user on DeepSeek V4.1 Flash 552B; H100 posts 2507 tok/s/chip for $0.13. GB200 NVL72 is 295% cheaper per token; GB200 NVL72 delivers 528% more tok/s/chip.
Throughput at 243 tok/s/user on DeepSeek V4.1 Flash 552B: GB200 NVL72 hits 13306 tok/s/chip, H100 hits 2481. Per-million costs land at $0.04 and $0.13 respectively. GB200 NVL72 is 237% cheaper per token; GB200 NVL72 delivers 436% more tok/s/chip.
GB200 NVL72 / H100 on DeepSeek V4.1 Flash 552B at 259 tok/s/user: 11034 / 2466 tok/s/chip, $0.05 / $0.13 per million tokens. GB200 NVL72 is 181% cheaper per token; GB200 NVL72 delivers 347% more tok/s/chip. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB200 NVL72:15745.3H100:2507.3 | GB200 NVL72:13305.7H100:2481.2 | GB200 NVL72:11034.1H100:2465.9 |
| Cost ($/M tok) | GB200 NVL72:$0.033H100:$0.130 | GB200 NVL72:$0.039H100:$0.131 | GB200 NVL72:$0.047H100:$0.132 |
| tok/s/MW | GB200 NVL72:8419943H100:1830166 | GB200 NVL72:7115352H100:1811072 | GB200 NVL72:5900578H100:1799936 |
| Concurrency | GB200 NVL72:~8H100:~2 | GB200 NVL72:~7H100:~1 | GB200 NVL72:~6H100:~1 |
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.
