Kimi K2.5/K2.6 1T — B300 vs H200 Performance per Dollar
Cost per million tokens of B300 (NVIDIA Blackwell) versus H200 (NVIDIA Hopper) on Kimi K2.5/K2.6 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.
Near the low end of the 34–106 tok/s/user interactivity band — at 51 tok/s/user — B300 runs $1.41 per million tokens on Kimi K2.5/K2.6 1T while H200 runs $1.23. H200 is the cheaper choice by 14%.
On Kimi K2.5/K2.6 1T at 70 tok/s/user, the per-million math comes out to $1.99 for B300 and $1.77 for H200; H200 delivers 12% more output per dollar.
At 88 tok/s/user on Kimi K2.5/K2.6 1T, B300 costs $2.81 per million tokens; H200 costs $2.50. H200 is 12% more cost-efficient at this operating point. (Numbers reflect the default 1k/1k · int4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
GPU pricing (owning hyperscaler): B300 $2.34/GPU/hr · H200 $1.41/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 | B300:$1.412H200:$1.234 | B300:$1.995H200:$1.775 | B300:$2.809H200:$2.500 |
| Concurrency | B300:~32H200:~25 | B300:~13H200:~13 | B300:~5H200:~7 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.