Grok 4 · xAI, released 2025-07-09
o3 · OpenAI, released 2025-04-16
“Epoch AI, 'AI Benchmarking Hub'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/benchmarks' [online resource].” · epoch.ai/benchmarks · Licensed CC BY 4.0 · data last verified 2026-07-26 · Methodology · Data Sources
Measured on 6 of 9 tracked benchmarks for both models — Grok 4 leads on 3, o3 leads on 2, tied on 1.
| Benchmark | Grok 4 | o3 | Leads |
|---|---|---|---|
| Epoch Capabilities Index (ECI) | 147.0 | 147.1 | o3 |
| SWE-bench Verified | Not measured | 0.623 (medium) | Not measured on both |
| GPQA Diamond | 0.870 | 0.818 (high) | Grok 4 |
| MATH Level 5 | Not measured | 0.978 (high) | Not measured on both |
| OTIS Mock AIME 2024-2025 | 0.840 | 0.839 (high) | Grok 4 |
| FrontierMath | 0.197 | 0.187 (high) | Grok 4 |
| FrontierMath Tier 4 | 0.021 | 0.021 (high) | Tied |
| SimpleQA Verified | 0.479 | 0.530 (high) | o3 |
| Chess Puzzles | 0.280 | Not measured | Not measured on both |
Xentir