Qwen 3.5 397B-A17B — B200 vs B300
Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and B300 (NVIDIA Blackwell) on Qwen 3.5 397B-A17B. 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.
Throughput at 85 tok/s/user on Qwen 3.5 397B-A17B: B200 hits 2256 tok/s/GPU, B300 hits 2238. Per-million costs land at $0.24 and $0.29 respectively. B200 is 22% cheaper per token; throughput per GPU is essentially tied.
B200 / B300 on Qwen 3.5 397B-A17B at 140 tok/s/user: 939 / 1079 tok/s/GPU, $0.58 / $0.60 per million tokens. B200 is 5% cheaper per token; B300 delivers 15% more tok/s/GPU.
Toward the upper edge of the 30–249 tok/s/user interactivity band, at 195 tok/s/user on Qwen 3.5 397B-A17B: B200 runs 478 tok/s/GPU at $1.09/M tokens, B300 runs 697 at $0.93/M. B300 is 17% cheaper per token; B300 delivers 46% more tok/s/GPU. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | B200:2256.4B300:2237.7 | B200:939.2B300:1079.4 | B200:477.5B300:697.2 |
| Cost ($/M tok) | B200:$0.239B300:$0.292 | B200:$0.577B300:$0.603 | B200:$1.087B300:$0.931 |
| tok/s/MW | B200:1039836B300:1031184 | B200:432834B300:497438 | B200:220059B300:321297 |
| Concurrency | B200:~58B300:~54 | B200:~14B300:~16 | B200:~5B300:~8 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.