Head-to-head AI inference benchmark comparison of B300 (NVIDIA Blackwell) and GB200 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
B300 / GB200 NVL72 on GLM 5.3 744B at 129 tok/s/user: 12614 / 14602 tok/s/chip, $0.05 / $0.04 per million tokens. GB200 NVL72 is 41% cheaper per token; GB200 NVL72 delivers 16% more tok/s/chip.
Around the middle of the 81β273 tok/s/user interactivity band, at 177 tok/s/user on GLM 5.3 744B: B300 runs 7500 tok/s/chip at $0.08/M tokens, GB200 NVL72 runs 10626 at $0.05/M. GB200 NVL72 is 72% cheaper per token; GB200 NVL72 delivers 42% more tok/s/chip.
Setting 225 tok/s/user as the target on GLM 5.3 744B, B300 produces 5358 tok/s/chip ($0.12 per million tokens) and GB200 NVL72 produces 5814 ($0.09). GB200 NVL72 is 32% cheaper per token; GB200 NVL72 delivers 9% more tok/s/chip. (Numbers reflect the default agentic-traces Β· fp4 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B300:12614.5GB200 NVL72:14601.7 | B300:7499.7GB200 NVL72:10626.1 | B300:5358.0GB200 NVL72:5813.6 |
| Cost ($/M tok) | B300:$0.050GB200 NVL72:$0.035 | B300:$0.084GB200 NVL72:$0.049 | B300:$0.117GB200 NVL72:$0.089 |
| tok/s/MW | B300:6639188GB200 NVL72:7808405 | B300:3947220GB200 NVL72:5682411 | B300:2820005GB200 NVL72:3108884 |
| Concurrency | B300:~16GB200 NVL72:~83 | B300:~11GB200 NVL72:~48 | B300:~6GB200 NVL72:~6 |
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.
