DeepSeek R1 — B200 vs Jalapeño
Head-to-head AI inference benchmark comparison of B200 (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.
B200 / Jalapeño on DeepSeek R1 at 124 tok/s/user: 3046 / 11215 tok/s/chip, $0.16 / $0.03 per million tokens. Jalapeño is 402% cheaper per token; Jalapeño delivers 268% more tok/s/chip.
Around the middle of the 66–300 tok/s/user interactivity band, at 183 tok/s/user on DeepSeek R1: B200 runs 1839 tok/s/chip at $0.26/M tokens, Jalapeño runs 7781 at $0.05/M. Jalapeño is 476% cheaper per token; Jalapeño delivers 323% more tok/s/chip.
Setting 242 tok/s/user as the target on DeepSeek R1, B200 produces 1240 tok/s/chip ($0.39 per million tokens) and Jalapeño produces 2562 ($0.14). Jalapeño is 182% cheaper per token; Jalapeño delivers 107% more tok/s/chip. (Numbers reflect the default 8k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B200:3045.9Jalapeño:11215.4 | B200:1839.3Jalapeño:7780.5 | B200:1239.6Jalapeño:2562.0 |
| Cost ($/M tok) | B200:$0.158Jalapeño:$0.031 | B200:$0.261Jalapeño:$0.045 | B200:$0.388Jalapeño:$0.138 |
| tok/s/MW | B200:1781252Jalapeño:9969252 | B200:1075605Jalapeño:6916013 | B200:724919Jalapeño:2277347 |
| Concurrency | B200:~21Jalapeño:~685 | B200:~15Jalapeño:~255 | B200:~15Jalapeño:~47 |