GLM 5/5.1 · Performance per Dollar

GLM 5/5.1 — H200 vs MI325X Performance per Dollar

Cost per million tokens of H200 (NVIDIA Hopper) versus MI325X (AMD CDNA 3) on GLM 5/5.1. 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.

At 20 tok/s/user on GLM 5/5.1, H200 costs $0.40 per million tokens; MI325X costs $1.08. H200 delivers 172% more tokens per dollar at this operating point.

H200 edges MI325X at 23 tok/s/user on GLM 5/5.1 — $0.40 per million tokens versus $1.40, a 72% lower cost per token.

Push GLM 5/5.1 to 26 tok/s/user and H200 lands at $0.40 per million tokens against MI325X's $1.92 — H200 costs 79% less. (Numbers reflect the default 8k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)

Chip pricing (owning hyperscaler): H200 $1.22/chip/hr · MI325X $1.10/chip/hr. Source: SemiAnalysis Market July 2026 Pricing Surveys & AI Cloud TCO Model.

View full latency and throughput comparison →

MI325X vs H200: full specs & benchmarks across models →

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)
Dollars per Million Tokens
H200$0.40MI325X$1.08
H200$0.40MI325X$1.40
H200$0.40MI325X$1.92
Concurrency
H200~32MI325X~13
H200~32MI325X~9
H200~31MI325X~6
GLM 5/5.1: H200 versus MI325X cost per million tokens at matched interactivity levels
H200 versus MI325X cost per million tokens for this comparison's canonical default workload. Lower cost indicates better performance per dollar.
Open full-size image ↗