Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and B300 (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 — 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 →
B200 posts 7540 tok/s/chip for $0.06 per million tokens at 169 tok/s/user on GLM 5.3 744B; B300 posts 7450 tok/s/chip for $0.08. B200 is 32% cheaper per token; B200 delivers 1% more tok/s/chip.
Throughput at 213 tok/s/user on GLM 5.3 744B: B200 hits 6240 tok/s/chip, B300 hits 5283. Per-million costs land at $0.08 and $0.12 respectively. B200 is 54% cheaper per token; B200 delivers 18% more tok/s/chip.
B200 / B300 on GLM 5.3 744B at 257 tok/s/user: 4006 / 4159 tok/s/chip, $0.12 / $0.15 per million tokens. B200 is 26% cheaper per token; B300 delivers 4% more tok/s/chip. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B200:7540.5B300:7449.9 | B200:6239.5B300:5283.4 | B200:4006.1B300:4159.3 |
| Cost ($/M tok) | B200:$0.064B300:$0.084 | B200:$0.077B300:$0.119 | B200:$0.120B300:$0.151 |
| tok/s/MW | B200:4409626B300:3921012 | B200:3648855B300:2780745 | B200:2342727B300:2189099 |
| Concurrency | B200:~10B300:~9 | B200:~7B300:~6 | B200:~4B300:~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.
