Grok 4 vs o3
Côte à côte. Chaque métrique. Chaque benchmark.
| Type | Grok 4 | o3 |
|---|---|---|
| Provider |  xAI |  OpenAI |
| score moyen | 54.8 | 55.2 |
| Prix d'entrée | $3.00 | $2.00 |
| Prix de sortie | $15.00 | $8.00 |
| Fenêtre de contexte | 256K tokens (~128 books) | 200K tokens (~100 books) |
| Sorti le | 2025-07-09 | 2025-04-16 |
| Code source ouvert | Proprietary | Proprietary |
Scores de benchmark
19 benchmarks · Grok 4: 9, o3: 9
| Benchmark | Catégorie | Grok 4 | o3 |
|---|---|---|---|
| Aider polyglot | coding | 79.6 | 81.3 |
| ARC-AGI | reasoning | 66.7 | 60.8 |
| ARC-AGI-2 | reasoning | 16.0 | 6.5 |
| DeepResearch Bench | knowledge | 47.9 | 46.6 |
| Fiction.LiveBench | knowledge | 94.4 | 88.9 |
| FrontierMath-2025-02-28-Private | math | 19.7 | 18.7 |
| FrontierMath-Tier-4-2025-07-01-Private | math | 2.1 | 2.1 |
| GeoBench | knowledge | 45.0 | 74.0 |
| GPQA diamond | knowledge | 82.7 | 75.8 |
| HELM — GPQA | knowledge | 72.6 | 75.3 |
| HELM — IFEval | language | 94.9 | 86.9 |
| HELM — MMLU-Pro | knowledge | 85.1 | 85.9 |
| HELM — Omni-MATH | math | 60.3 | 71.4 |
| HELM — WildBench | reasoning | 79.7 | 86.1 |
| Lech Mazur Writing | knowledge | 80.7 | 83.9 |
| OTIS Mock AIME 2024-2025 | math | 84.0 | 83.9 |
| SimpleBench | reasoning | 52.6 | 43.7 |
| SimpleQA Verified | knowledge | 47.9 | 53.0 |
| WeirdML | coding | 45.7 | 52.4 |