Kimi K2.5/K2.6/K2.7-Code 1T — B200 vs B300 Performance per Dollar
Cost per million tokens of B200 (NVIDIA Blackwell) versus B300 (NVIDIA Blackwell) on Kimi K2.5/K2.6/K2.7-Code 1T. 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: $0.11 per million tokens. B300: $0.21. Both at 59 tok/s/user on Kimi K2.5/K2.6/K2.7-Code 1T, with B200 87% cheaper.
Around the middle of the 23–167 tok/s/user interactivity band — at 95 tok/s/user — B200 runs $0.26 per million tokens on Kimi K2.5/K2.6/K2.7-Code 1T while B300 runs $0.24. B300 is the cheaper choice by 8%.
On Kimi K2.5/K2.6/K2.7-Code 1T at 131 tok/s/user, the per-million math comes out to $0.56 for B200 and $0.50 for B300; B300 delivers 13% more output per dollar. (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.)
GPU pricing (owning hyperscaler): B200 $1.95/GPU/hr · B300 $2.34/GPU/hr. Source: SemiAnalysis Market August 2025 Pricing Surveys & AI Cloud TCO Model.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Dollar per Million Tokens | B200:$0.114B300:$0.213 | B200:$0.257B300:$0.237 | B200:$0.562B300:$0.498 |
| Concurrency | B200:~222B300:~25 | B200:~95B300:~14 | B200:~30B300:~5 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.