GPT-5 mini (High) vs Granite 4.2 30B
Which is better for the money: GPT-5 mini (High) by OpenAI or Granite 4.2 30B by IBM? 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 mini (High) scores 4.1 points higher (16.8 vs 12.7) but costs 1.7× as much to run the benchmark ($189 vs $111; both mid-range class). Pick GPT-5 mini (High) when the extra intelligence is worth the premium, otherwise Granite 4.2 30B.
| Metric | GPT-5 mini (High) | Granite 4.2 30B | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 16.8 | 12.7 | −24.4% Granite 4.2 30B is 24.4% less intelligent |
| Total AA cost (lower is better) | $188.66 | $110.83 | −41.3% Granite 4.2 30B is 41.3% cheaper |
| End-to-end time (lower is better) | 58.74s | 51.51s | −12.3% Granite 4.2 30B is 12.3% faster |
| Price per 1M tokens (in / out) | $0.25 / $2.00 | $0.16 / $0.65 | |
| Input modalities | Text, Image | Text |
Granite 4.2 30B vs GPT-5 mini (High): 24.4% less intelligence, 41.3% less cost to run the AA benchmark, and 12.3% faster end-to-end.
Today's Pareto frontier
Against the frontier
- GPT-5 mini (High) is off the frontier: at the same or lower cost, Claude Haiku 5.5 (Xhigh) scores 41.20 for $158.
- Granite 4.2 30B is off the frontier: at the same or lower cost, MiMo-V2.6-Flash scores 37.90 for $110.
GPT-5 mini (High)Granite 4.2 30BToday's Pareto frontierOpen-weights 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-mini/granite-4-2-30b · Markdown: /api/v1/compare/gpt-5-mini/granite-4-2-30b?format=md · llms.txt · agent skill