Head-to-head AI inference benchmark comparison of H100 (NVIDIA Hopper) and H200 (NVIDIA Hopper) on Qwen 3.5 397B-A17B. 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 →
At 96 tok/s/user interactivity on Qwen 3.5 397B-A17B, H100 delivers 5072 tok/s/chip at $0.06 per million tokens; H200 delivers 8293 tok/s/chip at $0.04. H200 is 57% cheaper per token; H200 delivers 63% more tok/s/chip at this point.
H100 posts 3150 tok/s/chip for $0.10 per million tokens at 140 tok/s/user on Qwen 3.5 397B-A17B; H200 posts 5549 tok/s/chip for $0.06. H200 is 69% cheaper per token; H200 delivers 76% more tok/s/chip.
Throughput at 185 tok/s/user on Qwen 3.5 397B-A17B: H100 hits 2379 tok/s/chip, H200 hits 3131. Per-million costs land at $0.14 and $0.11 respectively. H200 is 26% cheaper per token; H200 delivers 32% more tok/s/chip. (Numbers reflect the default agentic-traces · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | H100:5072.3H200:8292.6 | H100:3149.9H200:5548.9 | H100:2379.1H200:3131.4 |
| Cost ($/M tok) | H100:$0.064H200:$0.041 | H100:$0.103H200:$0.061 | H100:$0.137H200:$0.108 |
| tok/s/MW | H100:3702377H200:6052985 | H100:2299197H200:4050279 | H100:1736572H200:2285669 |
| Concurrency | H100:~10H200:~13 | H100:~5H200:~9 | H100:~4H200:~4 |
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
TCO $/chip/hr:
Source: SemiAnalysis Market July 2026 Pricing Surveys & AI Cloud TCO Model
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate cost per million tokens per decode chip or per prefill chip, rather than per total chip count. This makes direct token cost comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate input throughput per decode chip or per prefill chip, rather than per total chip count. This makes direct input throughput comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate output throughput per decode chip or per prefill chip, rather than per total chip count. This makes direct output throughput comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate power per decode chip or per prefill chip, rather than per total chip count. This makes direct power comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate Joules per decode chip or per prefill chip, rather than per total chip count. This makes direct Joules per token comparison with aggregated configs not an apples-to-apples comparison.
No measurements to plot for this selection. Review the benchmark controls above or adjust quick filters.
Shift+Scroll to zoom • Drag to pan • Double-click to reset • Click a point to pin tooltip
Qwen3.5 397B 8K / 1K Total Tokens per $1 TCO vs. Interactivity