Kimi K3 2.8T — B200 vs GB200 NVL72
Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and GB200 NVL72 (NVIDIA Blackwell) on Kimi K3 2.8T. 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.
AgentX replays real coding-agent sessions rather than fixed-length prompts, so context grows turn over turn and most of each request is served from cache instead of being recomputed. That turns the comparison into a systems question: KV transfer between nodes, prefix-aware routing, and cache capacity all move the curve alongside raw chip throughput. Fixed-sequence workloads stay the clean baseline for kernel and silicon performance, so the two scenarios answer different questions about the same hardware. Learn more about AgentX →
B200 posts 4478 tok/s/chip for $0.11 per million tokens at 51 tok/s/user on Kimi K3 2.8T; GB200 NVL72 posts 4502 tok/s/chip for $0.11. B200 is 7% cheaper per token; throughput per chip is essentially tied.
Throughput at 100 tok/s/user on Kimi K3 2.8T: B200 hits 2147 tok/s/chip, GB200 NVL72 hits 2273. Per-million costs land at $0.22 and $0.23 respectively. B200 is 2% cheaper per token; GB200 NVL72 delivers 6% more tok/s/chip.
B200 / GB200 NVL72 on Kimi K3 2.8T at 149 tok/s/user: 1378 / 1282 tok/s/chip, $0.35 / $0.40 per million tokens. B200 is 16% cheaper per token; B200 delivers 8% more tok/s/chip. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B200:4478.1GB200 NVL72:4502.2 | B200:2146.8GB200 NVL72:2273.2 | B200:1378.1GB200 NVL72:1281.6 |
| Cost ($/M tok) | B200:$0.107GB200 NVL72:$0.115 | B200:$0.224GB200 NVL72:$0.227 | B200:$0.349GB200 NVL72:$0.403 |
| tok/s/MW | B200:2618770GB200 NVL72:2407606 | B200:1255456GB200 NVL72:1215613 | B200:805892GB200 NVL72:685337 |
| Concurrency | B200:~16GB200 NVL72:~16 | B200:~6GB200 NVL72:~15 | B200:~3GB200 NVL72:~3 |