Kimi K3 2.8T · Chip comparison
Kimi K3 2.8T — B200 vs GB300 NVL72
Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and GB300 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 →
Near the low end of the 35–186 tok/s/user interactivity band, at 72 tok/s/user on Kimi K3 2.8T: B200 runs 4,015 tok/s/chip at $0.12 per million tokens, GB300 NVL72 runs 19,064 at $0.034. GB300 NVL72 is 72% cheaper per token; GB300 NVL72 delivers 4.7× the tok/s/chip.
Setting 110 tok/s/user as the target on Kimi K3 2.8T, B200 produces 2,363 tok/s/chip ($0.20 per million tokens) and GB300 NVL72 produces 11,604 ($0.055). GB300 NVL72 is 73% cheaper per token; GB300 NVL72 delivers 4.9× the tok/s/chip.
At 148 tok/s/user interactivity on Kimi K3 2.8T, B200 delivers 1,687 tok/s/chip at $0.28 per million tokens; GB300 NVL72 delivers 10,183 tok/s/chip at $0.063. GB300 NVL72 is 78% cheaper per token; GB300 NVL72 delivers 6.0× the tok/s/chip at this point. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B2004,015GB300 NVL7219,064 | B2002,363GB300 NVL7211,604 | B2001,687GB300 NVL7210,183 |
| Cost ($/M tok) | B200$0.12GB300 NVL72$0.034 | B200$0.20GB300 NVL72$0.055 | B200$0.28GB300 NVL72$0.063 |
| tok/s/MW | B2002,347,748GB300 NVL728,992,465 | B2001,381,721GB300 NVL725,473,779 | B200986,420GB300 NVL724,803,511 |
| Concurrency | B200~13GB300 NVL72~220 | B200~6GB300 NVL72~47 | B200~4GB300 NVL72~35 |