gpt-oss 120B — B200 vs H100 Performance per Dollar
Cost per million tokens of B200 (NVIDIA Blackwell) versus H100 (NVIDIA Hopper) on gpt-oss 120B. Owning-hyperscaler TCO normalized by output tokens — performance per dollar across LLM workloads. Pick the more cost-efficient SKU at every target interactivity level. Use the chart controls below to switch sequences, precisions, and metrics — same interactions as the main inference chart.
B200: $0.03 per million tokens. H100: $0.13. Both at 117 tok/s/user on gpt-oss 120B, with B200 335% cheaper.
Around the middle of the 67–266 tok/s/user interactivity band — at 166 tok/s/user — B200 runs $0.05 per million tokens on gpt-oss 120B while H100 runs $0.24. B200 is the cheaper choice by 354%.
On gpt-oss 120B at 216 tok/s/user, the per-million math comes out to $0.08 for B200 and $0.44 for H100; B200 delivers 419% more output per dollar. (Numbers reflect the default 1k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Chip pricing (owning hyperscaler): B200 $1.73/chip/hr · H100 $1.17/chip/hr. Source: SemiAnalysis Market July 2026 Pricing Surveys & AI Cloud TCO Model.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Dollar per Million Tokens | B200:$0.029H100:$0.125 | B200:$0.052H100:$0.236 | B200:$0.084H100:$0.438 |
| Concurrency | B200:~154H100:~64 | B200:~64H100:~17 | B200:~61H100:~8 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.