o1 vs o3
Côte à côte. Chaque métrique. Chaque benchmark.
| Type | o1 | o3 |
|---|---|---|
| Provider |  OpenAI | OpenAI |
| score moyen | 56.4 | 55.2 |
| Prix d'entrée | $15.00 | $2.00 |
| Prix de sortie | $60.00 | $8.00 |
| Fenêtre de contexte | 200K tokens (~100 books) | 200K tokens (~100 books) |
| Sorti le | 2024-12-17 | 2025-04-16 |
| Code source ouvert | Proprietary | Proprietary |
Scores de benchmark
13 benchmarks · o1: 1, o3: 12
| Benchmark | Catégorie | o1 | o3 |
|---|---|---|---|
| Aider polyglot | coding | 61.7 | 81.3 |
| ARC-AGI | reasoning | 30.7 | 60.8 |
| CadEval | coding | 56.0 | 74.0 |
| Fiction.LiveBench | knowledge | 83.3 | 88.9 |
| FrontierMath-2025-02-28-Private | math | 9.3 | 18.7 |
| GeoBench | knowledge | 80.0 | 74.0 |
| GPQA diamond | knowledge | 69.0 | 75.8 |
| Lech Mazur Writing | knowledge | 70.2 | 83.9 |
| MATH level 5 | math | 94.7 | 97.8 |
| OTIS Mock AIME 2024-2025 | math | 73.3 | 83.9 |
| SimpleBench | reasoning | 28.1 | 43.7 |
| VPCT | knowledge | 5.5 | 28.0 |
| WeirdML | coding | 43.8 | 52.4 |