DeepSeek V4.1 Flash 552B — H100 vs MI300X Performance per Dollar
Cost per million tokens of H100 (NVIDIA Hopper) versus MI300X (AMD CDNA 3) on DeepSeek V4.1 Flash 552B. Large-hyperscaler-volume ownership TCO normalized by total tokens — performance per dollar across LLM workloads. Pick the more cost-efficient SKU at every target interactivity level. 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 →
H100 edges MI300X at 68 tok/s/user on DeepSeek V4.1 Flash 552B — $0.02 per million tokens versus $0.02, a 35% cost-per-token gap.
Push DeepSeek V4.1 Flash 552B to 113 tok/s/user and H100 lands at $0.02 per million tokens against MI300X's $0.04 — H100 pulls ahead by 105%.
H100: $0.03 per million tokens. MI300X: $0.07. Both at 159 tok/s/user on DeepSeek V4.1 Flash 552B, with H100 91% cheaper. (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.)
Chip pricing (owning hyperscaler): H100 $1.17/chip/hr · MI300X $0.95/chip/hr. Source: SemiAnalysis Market July 2026 Pricing Surveys & AI Cloud TCO Model.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Dollar per Million Tokens | H100:$0.016MI300X:$0.022 | H100:$0.018MI300X:$0.038 | H100:$0.034MI300X:$0.065 |
| Concurrency | H100:~24MI300X:~9 | H100:~21MI300X:~5 | H100:~11MI300X:~1 |
Inference Performance
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.