Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and GB300 NVL72 (NVIDIA Blackwell) on GLM 5.3 744B. 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 150 tok/s/user as the target on GLM 5.3 744B, B200 produces 12589 tok/s/chip ($0.04 per million tokens) and GB300 NVL72 produces 15953 ($0.04). B200 is 5% cheaper per token; GB300 NVL72 delivers 27% more tok/s/chip.
At 208 tok/s/user interactivity on GLM 5.3 744B, B200 delivers 9014 tok/s/chip at $0.05 per million tokens; GB300 NVL72 delivers 10308 tok/s/chip at $0.06. B200 is 17% cheaper per token; GB300 NVL72 delivers 14% more tok/s/chip at this point.
B200 posts 6642 tok/s/chip for $0.07 per million tokens at 266 tok/s/user on GLM 5.3 744B; GB300 NVL72 posts 4008 tok/s/chip for $0.16. B200 is 121% cheaper per token; B200 delivers 66% more tok/s/chip. (Numbers reflect the default agentic-traces Β· fp4 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B200:12589.2GB300 NVL72:15953.3 | B200:9013.8GB300 NVL72:10307.9 | B200:6642.5GB300 NVL72:4007.6 |
| Cost ($/M tok) | B200:$0.038GB300 NVL72:$0.040 | B200:$0.053GB300 NVL72:$0.062 | B200:$0.072GB300 NVL72:$0.160 |
| tok/s/MW | B200:7362093GB300 NVL72:7525120 | B200:5271241GB300 NVL72:4862225 | B200:3884491GB300 NVL72:1890356 |
| Concurrency | B200:~9GB300 NVL72:~56 | B200:~5GB300 NVL72:~21 | B200:~2GB300 NVL72:~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.
