DeepSeek R1 — B200 vs GB300 NVL72 Performance per Dollar
Cost per million tokens of B200 (NVIDIA Blackwell) versus GB300 NVL72 (NVIDIA Blackwell) on DeepSeek R1. 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.
At 70 tok/s/user on DeepSeek R1, B200 costs $0.52 per million tokens; GB300 NVL72 costs $0.14. GB300 NVL72 is 279% more cost-efficient at this operating point.
GB300 NVL72 edges B200 at 127 tok/s/user on DeepSeek R1 — $0.77 per million tokens versus $1.97, a 157% cost-per-token gap.
Push DeepSeek R1 to 183 tok/s/user and B200 lands at $3.40 per million tokens against GB300 NVL72's $6.18 — B200 pulls ahead by 82%. (Numbers reflect the default 1k/1k · fp8 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 · GB300 NVL72 $2.65/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:$0.517GB300 NVL72:$0.136 | B200:$1.973GB300 NVL72:$0.767 | B200:$3.401GB300 NVL72:$6.176 |
| Concurrency | B200:~62GB300 NVL72:~959 | B200:~8GB300 NVL72:~196 | B200:~23GB300 NVL72:~20 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.