Head-to-head AI inference benchmark comparison of GB300 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 →
Near the low end of the 3–275 tok/s/user interactivity band, at 71 tok/s/user on DeepSeek V4.1 Flash 552B: GB300 NVL72 runs 10733 tok/s/chip at $0.06/M tokens, H100 runs 3829 at $0.08/M. GB300 NVL72 is 42% cheaper per token; GB300 NVL72 delivers 180% more tok/s/chip.
Setting 139 tok/s/user as the target on DeepSeek V4.1 Flash 552B, GB300 NVL72 produces 10733 tok/s/chip ($0.06 per million tokens) and H100 produces 3829 ($0.08). GB300 NVL72 is 42% cheaper per token; GB300 NVL72 delivers 180% more tok/s/chip.
At 207 tok/s/user interactivity on DeepSeek V4.1 Flash 552B, GB300 NVL72 delivers 10733 tok/s/chip at $0.06 per million tokens; H100 delivers 3829 tok/s/chip at $0.08. GB300 NVL72 is 42% cheaper per token; GB300 NVL72 delivers 180% more tok/s/chip at this point. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB300 NVL72:10733.4H100:3828.8 | GB300 NVL72:10733.4H100:3828.8 | GB300 NVL72:10733.4H100:3828.8 |
| Cost ($/M tok) | GB300 NVL72:$0.060H100:$0.085 | GB300 NVL72:$0.060H100:$0.085 | GB300 NVL72:$0.060H100:$0.085 |
| tok/s/MW | GB300 NVL72:5062944H100:2794727 | GB300 NVL72:5062944H100:2794727 | GB300 NVL72:5062944H100:2794727 |
| Concurrency | GB300 NVL72:~16H100:~4 | GB300 NVL72:~16H100:~4 | GB300 NVL72:~16H100:~4 |
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.
