Head-to-head AI inference benchmark comparison of B300 (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 — 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 27–392 tok/s/user interactivity band, at 118 tok/s/user on DeepSeek V4.1 Flash 552B: B300 runs 58064 tok/s/chip at $0.01/M tokens, GB300 NVL72 runs 66547 at $0.01/M. GB300 NVL72 is 12% cheaper per token; GB300 NVL72 delivers 15% more tok/s/chip.
Setting 210 tok/s/user as the target on DeepSeek V4.1 Flash 552B, B300 produces 35519 tok/s/chip ($0.02 per million tokens) and GB300 NVL72 produces 29940 ($0.02). B300 is 21% cheaper per token; B300 delivers 19% more tok/s/chip.
At 301 tok/s/user interactivity on DeepSeek V4.1 Flash 552B, B300 delivers 20199 tok/s/chip at $0.03 per million tokens; GB300 NVL72 delivers 19743 tok/s/chip at $0.03. B300 is 5% cheaper per token; B300 delivers 2% more tok/s/chip at this point. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B300:58064.4GB300 NVL72:66547.5 | B300:35519.1GB300 NVL72:29939.9 | B300:20198.9GB300 NVL72:19742.7 |
| Cost ($/M tok) | B300:$0.011GB300 NVL72:$0.010 | B300:$0.018GB300 NVL72:$0.021 | B300:$0.031GB300 NVL72:$0.033 |
| tok/s/MW | B300:30560235GB300 NVL72:31390310 | B300:18694237GB300 NVL72:14122582 | B300:10631017GB300 NVL72:9312572 |
| Concurrency | B300:~42GB300 NVL72:~21 | B300:~10GB300 NVL72:~8 | B300:~5GB300 NVL72:~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.
