o4 Mini
von OpenAI · Veroeffentlicht 2025-04-16
53.2
Durchschn. Score
$1.10/1M
Eingabepreis
$4.40/1M
Ausgabepreis
200K tokens (~100 books)
Kontextfenster
multimodal
Typ
Tested on 26 benchmarks with 53.2% average. Top scores: MATH level 5 (97.8%), HELM — IFEval (92.9%), HELM — WildBench (85.4%).
Benchmark-Ergebnisse
| Benchmark | Kategorie | Score | Bar |
|---|---|---|---|
| MATH level 5 | math | 97.8 | |
| HELM — IFEval | language | 92.9 | |
| HELM — WildBench | reasoning | 85.4 | |
| HELM — MMLU-Pro | knowledge | 82.0 | |
| OTIS Mock AIME 2024-2025 | math | 81.7 | |
| Fiction.LiveBench | knowledge | 77.8 | |
| Lech Mazur Writing | knowledge | 75.0 | |
| HELM — GPQA | knowledge | 73.5 | |
| GPQA diamond | knowledge | 72.8 | |
| Aider polyglot | coding | 72.0 | |
| HELM — Omni-MATH | math | 72.0 | |
| GeoBench | knowledge | 64.0 | |
| CadEval | coding | 62.0 | |
| ARC-AGI | reasoning | 58.7 | |
| WeirdML | coding | 52.6 | |
| VISTA | knowledge | 51.8 | |
| SWE-Bench Verified (Bash Only) | coding | 45.0 | |
| VPCT | knowledge | 36.3 | |
| SimpleBench | reasoning | 26.4 | |
| Chess Puzzles | knowledge | 26.0 | |
| FrontierMath-2025-02-28-Private | math | 24.8 | |
| SimpleQA Verified | knowledge | 23.9 | |
| HLE | knowledge | 13.9 | |
| FrontierMath-Tier-4-2025-07-01-Private | math | 6.3 | |
| ARC-AGI-2 | reasoning | 6.1 | |
| GSO-Bench | coding | 3.6 |