Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and GB300 NVL72 (NVIDIA Blackwell) on DeepSeek V4.1 Flash 552B. 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
Setting 200 tok/s/user as the target on DeepSeek V4.1 Flash 552B, B200 produces 68999 tok/s/chip ($0.01 per million tokens) and GB300 NVL72 produces 616176 ($0.00). GB300 NVL72 is 569% cheaper per token; GB300 NVL72 delivers 793% more tok/s/chip.
At 327 tok/s/user interactivity on DeepSeek V4.1 Flash 552B, B200 delivers 28750 tok/s/chip at $0.02 per million tokens; GB300 NVL72 delivers 29266 tok/s/chip at $0.02. B200 is 31% cheaper per token; GB300 NVL72 delivers 2% more tok/s/chip at this point.
B200 posts 9876 tok/s/chip for $0.05 per million tokens at 454 tok/s/user on DeepSeek V4.1 Flash 552B; GB300 NVL72 posts 24738 tok/s/chip for $0.03. GB300 NVL72 is 88% cheaper per token; GB300 NVL72 delivers 150% more tok/s/chip. (Numbers reflect the default agentic-traces Β· fp4 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B200:68998.7GB300 NVL72:616176.2 | B200:28750.2GB300 NVL72:29266.4 | B200:9875.7GB300 NVL72:24738.3 |
| Cost ($/M tok) | B200:$0.007GB300 NVL72:$0.001 | B200:$0.017GB300 NVL72:$0.022 | B200:$0.049GB300 NVL72:$0.026 |
| tok/s/MW | B200:40350125GB300 NVL72:290649150 | B200:16812964GB300 NVL72:13804912 | B200:5775245GB300 NVL72:11668999 |
| Concurrency | B200:~43GB300 NVL72:~2942 | B200:~16GB300 NVL72:~16 | B200:~4GB300 NVL72:~14 |
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.
