GPT-5.4 mini (Xhigh) vs GPT-5 mini (High)
Which is better for the money: GPT-5.4 mini (Xhigh) by OpenAI or GPT-5 mini (High) by OpenAI? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-10-25.
Trade-off
Verdict
Depends on your budget
GPT-5.4 mini (Xhigh) scores 7.3 points higher (24.1 vs 16.8) but costs 8.1× as much to run the benchmark ($1524 vs $189; premium vs mid-range cost classes). They are not alternatives at one budget: pick GPT-5.4 mini (Xhigh) when you need intelligence above 16.8, and GPT-5 mini (High) when 16.8 is enough.
| Metric | GPT-5.4 mini (Xhigh) | GPT-5 mini (High) | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 24.1 | 16.8 | −30.3% GPT-5 mini (High) is 30.3% less intelligent |
| Total AA cost (lower is better) | $1,524.02 | $188.66 | −87.6% GPT-5 mini (High) is 87.6% cheaper |
| End-to-end time (lower is better) | 109.06s | 60.90s | −44.2% GPT-5 mini (High) is 44.2% faster |
| Price per 1M tokens (in / out) | $0.75 / $4.50 | $0.25 / $2.00 | |
| Input modalities | Text, Image | Text, Image |
GPT-5 mini (High) vs GPT-5.4 mini (Xhigh): 30.3% less intelligence, 87.6% less cost to run the AA benchmark, and 44.2% faster end-to-end.
Today's Pareto frontier
Against the frontier
- GPT-5.4 mini (Xhigh) is off the frontier: at the same or lower cost, GPT-6.1 Sol (Max) scores 51.80 for $1082.
- GPT-5 mini (High) is off the frontier: at the same or lower cost, Claude Haiku 5.5 (Xhigh) scores 41.20 for $158.
GPT-5.4 mini (Xhigh)GPT-5 mini (High)Today's Pareto frontierEvery other model
Matchup
Compare something else
Agents:
Accept: text/markdown or ?format=md returns this page as Markdown. JSON: /api/v1/compare/gpt-5-4-mini/gpt-5-mini · Markdown: /api/v1/compare/gpt-5-4-mini/gpt-5-mini?format=md · llms.txt · agent skill