Llama 3.3 70B — H100 vs H200
Head-to-head AI inference benchmark comparison of H100 (NVIDIA Hopper) and H200 (NVIDIA Hopper) on Llama 3.3 70B. 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.
Throughput at 53 tok/s/user on Llama 3.3 70B: H100 hits 1465 tok/s/GPU, H200 hits 3036. Per-million costs land at $0.25 and $0.13 respectively. H200 is 94% cheaper per token; H200 delivers 107% more tok/s/GPU.
H100 / H200 on Llama 3.3 70B at 72 tok/s/user: 773 / 2184 tok/s/GPU, $0.45 / $0.18 per million tokens. H200 is 153% cheaper per token; H200 delivers 183% more tok/s/GPU.
Toward the upper edge of the 35–109 tok/s/user interactivity band, at 91 tok/s/user on Llama 3.3 70B: H100 runs 304 tok/s/GPU at $1.19/M tokens, H200 runs 1480 at $0.26/M. H200 is 354% cheaper per token; H200 delivers 387% more tok/s/GPU. (Numbers reflect the default 1k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/gpu) | H100:1465.2H200:3035.7 | H100:772.9H200:2184.2 | H100:303.7H200:1480.1 |
| Cost ($/M tok) | H100:$0.249H200:$0.129 | H100:$0.454H200:$0.179 | H100:$1.194H200:$0.263 |
| tok/s/MW | H100:846932H200:1754733 | H100:446737H200:1262549 | H100:175535H200:855562 |
| Concurrency | H100:~64H200:~64 | H100:~50H200:~64 | H100:~14H200:~35 |
Inference Performance
Inference performance metrics across different models, hardware configurations, and serving parameters.