Kimi K3 2.8T · Chip comparison

Kimi K3 2.8T — B200 vs H200

Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and H200 (NVIDIA Hopper) 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 9048 tok/s/chip for $0.05 per million tokens at 3 tok/s/user on Kimi K3 2.8T; H200 posts 118 tok/s/chip for $2.86. B200 is 5289% cheaper per token; B200 delivers 7542% more tok/s/chip.

Throughput at 4 tok/s/user on Kimi K3 2.8T: B200 hits 9048 tok/s/chip, H200 hits 91. Per-million costs land at $0.05 and $3.74 respectively. B200 is 6941% cheaper per token; B200 delivers 9884% more tok/s/chip.

B200 / H200 on Kimi K3 2.8T at 5 tok/s/user: 9048 / 72 tok/s/chip, $0.05 / $4.68 per million tokens. B200 is 8715% cheaper per token; B200 delivers 12399% 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.)

View performance-per-dollar view →

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)
B200:9047.6H200:118.4
B200:9047.6H200:90.6
B200:9047.6H200:72.4
Cost ($/M tok)
B200:$0.053H200:$2.862
B200:$0.053H200:$3.740
B200:$0.053H200:$4.682
tok/s/MW
B200:5290990H200:86423
B200:5290990H200:66144
B200:5290990H200:52836
Concurrency
B200:~96H200:~7
B200:~96H200:~5
B200:~96H200:~3

Inference Performance

Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.