Kimi K2.5/K2.6/K2.7-Code 1T — B300 vs MI325X Performance per Dollar
Cost per million tokens of B300 (NVIDIA Blackwell) versus MI325X (AMD CDNA 3) 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.
B300 edges MI325X at 33 tok/s/user on Kimi K2.5/K2.6/K2.7-Code 1T — $0.88 per million tokens versus $2.40, a 173% cost-per-token gap.
Push Kimi K2.5/K2.6/K2.7-Code 1T to 39 tok/s/user and B300 lands at $1.04 per million tokens against MI325X's $3.27 — B300 pulls ahead by 214%.
B300: $1.18 per million tokens. MI325X: $4.51. Both at 45 tok/s/user on Kimi K2.5/K2.6/K2.7-Code 1T, with B300 282% cheaper. (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.)
Chip pricing (owning hyperscaler): B300 $2.26/chip/hr · MI325X $1.10/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 | B300:$0.879MI325X:$2.397 | B300:$1.041MI325X:$3.266 | B300:$1.182MI325X:$4.513 |
| Concurrency | B300:~47MI325X:~16 | B300:~32MI325X:~10 | B300:~22MI325X:~6 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.