DeepSeek R1 — B300 vs Jalapeño
Head-to-head AI inference benchmark comparison of B300 (NVIDIA Blackwell) and Jalapeño (OpenAI Jalapeño) on DeepSeek R1. Latency, throughput, and cost across LLM workloads. Use the chart controls below to switch sequences, precisions, and metrics — same interactions as the main inference chart.
At 129 tok/s/user interactivity on DeepSeek R1, B300 delivers 3142 tok/s/chip at $0.20 per million tokens; Jalapeño delivers 10931 tok/s/chip at $0.03. Jalapeño is 519% cheaper per token; Jalapeño delivers 248% more tok/s/chip at this point.
B300 posts 1402 tok/s/chip for $0.45 per million tokens at 192 tok/s/user on DeepSeek R1; Jalapeño posts 6809 tok/s/chip for $0.05. Jalapeño is 764% cheaper per token; Jalapeño delivers 386% more tok/s/chip.
Throughput at 255 tok/s/user on DeepSeek R1: B300 hits 822 tok/s/chip, Jalapeño hits 2122. Per-million costs land at $0.76 and $0.17 respectively. Jalapeño is 359% cheaper per token; Jalapeño delivers 158% more tok/s/chip. (Numbers reflect the default 8k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B300:3142.0Jalapeño:10930.5 | B300:1402.3Jalapeño:6809.2 | B300:822.2Jalapeño:2122.4 |
| Cost ($/M tok) | B300:$0.200Jalapeño:$0.032 | B300:$0.448Jalapeño:$0.052 | B300:$0.764Jalapeño:$0.166 |
| tok/s/MW | B300:1653695Jalapeño:9716032 | B300:738034Jalapeño:6052645 | B300:432748Jalapeño:1886572 |
| Concurrency | B300:~64Jalapeño:~632 | B300:~40Jalapeño:~203 | B300:~27Jalapeño:~37 |