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 136 tok/s/user as the target on GLM 5.3 744B, B200 produces 10120 tok/s/chip ($0.05 per million tokens) and GB300 NVL72 produces 16442 ($0.04). GB300 NVL72 is 22% cheaper per token; GB300 NVL72 delivers 62% more tok/s/chip.
At 188 tok/s/user interactivity on GLM 5.3 744B, B200 delivers 6982 tok/s/chip at $0.07 per million tokens; GB300 NVL72 delivers 8119 tok/s/chip at $0.08. B200 is 15% cheaper per token; GB300 NVL72 delivers 16% more tok/s/chip at this point.
B200 posts 4953 tok/s/chip for $0.10 per million tokens at 240 tok/s/user on GLM 5.3 744B; GB300 NVL72 posts 4925 tok/s/chip for $0.13. B200 is 34% cheaper per token; throughput per chip is essentially tied. (Numbers reflect the default agentic-traces Β· fp4 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B200:10120.0GB300 NVL72:16441.6 | B200:6982.2GB300 NVL72:8119.4 | B200:4953.1GB300 NVL72:4925.3 |
| Cost ($/M tok) | B200:$0.047GB300 NVL72:$0.039 | B200:$0.069GB300 NVL72:$0.079 | B200:$0.097GB300 NVL72:$0.130 |
| tok/s/MW | B200:5918117GB300 NVL72:7755450 | B200:4083186GB300 NVL72:3829898 | B200:2896573GB300 NVL72:2323250 |
| Concurrency | B200:~33GB300 NVL72:~48 | B200:~8GB300 NVL72:~45 | B200:~6GB300 NVL72:~9 |
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.
