Best Benchmark Score
o4-mini
59.0/100
Model Specifications
| Spec | o3 | o4-mini |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Release Date | Apr 2025 | Apr 2025 |
| Knowledge Cutoff | 2024-05-31 | 2024-05-31 |
| Parameters | Undisclosed | Undisclosed |
| Context Window | — | — |
| Max Output | — | — |
| Spec | o3 | o4-mini |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Release Date | Apr 2025 | Apr 2025 |
| Knowledge Cutoff | 2024-05-31 | 2024-05-31 |
| Parameters | Undisclosed | Undisclosed |
| Context Window | — | — |
| Max Output | — | — |
| Open Source | No | No |
| License | proprietary | proprietary |
| Tokenizer | — | — |
| Modality | — | — |
| Reasoning Model | No | No |
| Moderated | No | No |
| Price (per 1M tokens) | o3 | o4-mini |
|---|
| Feature | o3 | o4-mini |
|---|---|---|
| Text Input | ||
| Image Input (Vision) | ||
| Audio Input | ||
| Video Input | ||
| File Input | ||
| Image Output | ||
| Audio Output | ||
| Tool Use | ||
| Structured Output (JSON) | ||
| Streaming | ||
| Reasoning Tokens | ||
| Web Search | ||
| Temperature Control | ||
| Top-P Sampling | ||
| Stop Sequences | ||
| Seed (Reproducibility) | ||
| Log Probabilities |
| Benchmark | o3 | o4-mini |
|---|---|---|
| SWE-bench Verified | 69.1 | 68.1 |
| Benchmark | o3 | o4-mini |
|---|---|---|
| AIME 2025 | 86.4 | 92.7 |
| FrontierMath | 15.8 | — |
| Benchmark | o3 | o4-mini |
|---|---|---|
| GPQA Diamond | 83.3 | 81.4 |
| Humanity's Last Exam | 14.7 | 14.7 |
| ARC-AGI v2 | 6.5 | — |
| Benchmark | o3 | o4-mini |
|---|---|---|
| MMMU | 82.9 | 81.6 |
| MMMU-Pro | 76.4 | — |
| CharXiv Reasoning | 78.6 | 72 |
| Benchmark | o3 | o4-mini |
|---|---|---|
| BrowseComp | 49.7 | 51.5 |
| TAU-Bench Retail | — | 71.8 |