GLM 5/5.1 — GB200 NVL72 vs H200
Head-to-head AI inference benchmark comparison of GB200 NVL72 (NVIDIA Blackwell) and H200 (NVIDIA Hopper) on GLM 5/5.1. 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 67 tok/s/user on GLM 5/5.1, GB200 NVL72 delivers 9135 tok/s/chip at $0.06 per million tokens; H200 hasn't been benchmarked at this target.
GB200 NVL72 hits 3619 tok/s/chip for $0.14 per million tokens at 104 tok/s/user on GLM 5/5.1. No H200 data at this operating point.
GB200 NVL72: 630 tok/s/chip, $0.82 per million tokens at 141 tok/s/user on GLM 5/5.1. H200 is unmeasured here. (Numbers reflect the default 8k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB200 NVL72:9135.4H200:— | GB200 NVL72:3618.9H200:— | GB200 NVL72:629.8H200:— |
| Cost ($/M tok) | GB200 NVL72:$0.057H200:— | GB200 NVL72:$0.143H200:— | GB200 NVL72:$0.820H200:— |
| tok/s/MW | GB200 NVL72:4885238H200:— | GB200 NVL72:1935265H200:— | GB200 NVL72:336816H200:— |
| Concurrency | GB200 NVL72:~698H200:— | GB200 NVL72:~231H200:— | GB200 NVL72:~62H200:— |
Inference Performance
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.