Head-to-head AI inference benchmark comparison of GB300 NVL72 (NVIDIA Blackwell) and H200 (NVIDIA Hopper) on Muse Spark Open 816B. 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
At 56 tok/s/user interactivity on Muse Spark Open 816B, GB300 NVL72 delivers 22642 tok/s/chip at $0.03 per million tokens; H200 delivers 6098 tok/s/chip at $0.06. GB300 NVL72 is 96% cheaper per token; GB300 NVL72 delivers 271% more tok/s/chip at this point.
GB300 NVL72 posts 20454 tok/s/chip for $0.03 per million tokens at 99 tok/s/user on Muse Spark Open 816B; H200 posts 4532 tok/s/chip for $0.07. GB300 NVL72 is 138% cheaper per token; GB300 NVL72 delivers 351% more tok/s/chip.
Throughput at 143 tok/s/user on Muse Spark Open 816B: GB300 NVL72 hits 17738 tok/s/chip, H200 hits 2823. Per-million costs land at $0.04 and $0.12 respectively. GB300 NVL72 is 232% cheaper per token; GB300 NVL72 delivers 528% more tok/s/chip. (Numbers reflect the default agentic-traces Β· fp4 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB300 NVL72:22642.0H200:6097.9 | GB300 NVL72:20454.5H200:4531.5 | GB300 NVL72:17738.2H200:2822.7 |
| Cost ($/M tok) | GB300 NVL72:$0.028H200:$0.056 | GB300 NVL72:$0.031H200:$0.075 | GB300 NVL72:$0.036H200:$0.120 |
| tok/s/MW | GB300 NVL72:10680176H200:4451017 | GB300 NVL72:9648340H200:3307691 | GB300 NVL72:8367078H200:2060348 |
| Concurrency | GB300 NVL72:~20H200:~13 | GB300 NVL72:~16H200:~8 | GB300 NVL72:~13H200:~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.
