o3
Unknown SizeBy OpenAI • Released 2025-04-16
Capability Radar
Avg Score
64
Across all benchmarks
Participated
14
Benchmarks
Benchmark Performance
| Benchmark | Category | Score |
|---|---|---|
| MATH-500 | Reasoning | 99.2 |
| AIME 2025 | Reasoning | 88.3 |
| MMLU-Pro | Knowledge | 85.3 |
| GPQA Diamond | Knowledge | 82.7 |
| LiveCodeBench | Coding | 80.8 |
| 𝜏²-Bench Telecom | Reasoning Knowledge | 80.7 |
| IFBench | Agent | 71.4 |
| LCR | Long-Context Reasoning | 69.3 |
| SWE-bench (Bash Only) | Coding Agent | 58.4 |
| SciCode | Reasoning Knowledge | 41 |
| Artificial Analysis Coding Index | Coding | 38.4 |
| Artificial Analysis Intelligence Index | Knowledge | 38.3 |
| Terminal-Bench Hard | Agent Coding | 37.1 |
| HLE | Knowledge Multi-Modal | 20 |