gpt-oss 120B · Chip comparison
gpt-oss 120B — B200 vs GB200 NVL72
Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and GB200 NVL72 (NVIDIA Blackwell) on gpt-oss 120B. 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.
B200 posts 22,825 tok/s/chip for $0.021 per million tokens at 155 tok/s/user on gpt-oss 120B; GB200 NVL72 posts 33,534 tok/s/chip for $0.015. GB200 NVL72 is 27% cheaper per token; GB200 NVL72 delivers 47% more tok/s/chip.
Throughput at 243 tok/s/user on gpt-oss 120B: B200 hits 13,355 tok/s/chip, GB200 NVL72 hits 19,526. Costs per million tokens land at $0.036 and $0.026 respectively. GB200 NVL72 is 26% cheaper per token; GB200 NVL72 delivers 46% more tok/s/chip.
B200 / GB200 NVL72 on gpt-oss 120B at 330 tok/s/user: 6,328 / 8,905 tok/s/chip, $0.076 / $0.058 per million tokens. GB200 NVL72 is 24% cheaper per token; GB200 NVL72 delivers 41% more tok/s/chip. (Numbers reflect the default 8k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
GB200 NVL72 vs B200: full specs & benchmarks across models →
Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B20022,825GB200 NVL7233,534 | B20013,355GB200 NVL7219,526 | B2006,328GB200 NVL728,905 |
| Cost ($/M tok) | B200$0.021GB200 NVL72$0.015 | B200$0.036GB200 NVL72$0.026 | B200$0.076GB200 NVL72$0.058 |
| tok/s/MW | B20013,348,051GB200 NVL7217,932,795 | B2007,810,104GB200 NVL7210,441,533 | B2003,700,528GB200 NVL724,761,975 |
| Concurrency | B200~32GB200 NVL72~343 | B200~6GB200 NVL72~56 | B200~8GB200 NVL72~17 |