gpt-oss 120B · Chip comparison
gpt-oss 120B — GB200 NVL72 vs H100
Head-to-head AI inference benchmark comparison of GB200 NVL72 (NVIDIA Blackwell) and H100 (NVIDIA Hopper) 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.
At 113 tok/s/user interactivity on gpt-oss 120B, GB200 NVL72 delivers 42,816 tok/s/chip at $0.012 per million tokens; H100 delivers 5,280 tok/s/chip at $0.062. GB200 NVL72 is 80% cheaper per token; GB200 NVL72 delivers 8.1× the tok/s/chip at this point.
GB200 NVL72 posts 32,911 tok/s/chip for $0.016 per million tokens at 159 tok/s/user on gpt-oss 120B; H100 posts 3,672 tok/s/chip for $0.089. GB200 NVL72 is 82% cheaper per token; GB200 NVL72 delivers 9.0× the tok/s/chip.
Throughput at 205 tok/s/user on gpt-oss 120B: GB200 NVL72 hits 25,756 tok/s/chip, H100 hits 2,428. Costs per million tokens land at $0.020 and $0.13 respectively. GB200 NVL72 is 85% cheaper per token; GB200 NVL72 delivers 11× the tok/s/chip. (Numbers reflect the default 8k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB200 NVL7242,816H1005,280 | GB200 NVL7232,911H1003,672 | GB200 NVL7225,756H1002,428 |
| Cost ($/M tok) | GB200 NVL72$0.012H100$0.062 | GB200 NVL72$0.016H100$0.089 | GB200 NVL72$0.020H100$0.13 |
| tok/s/MW | GB200 NVL7222,896,080H1003,853,936 | GB200 NVL7217,599,435H1002,680,055 | GB200 NVL7213,773,170H1001,772,304 |
| Concurrency | GB200 NVL72~375H100~25 | GB200 NVL72~291H100~11 | GB200 NVL72~87H100~5 |