Scientific Evaluation on SciEval
87.5Accuracyo4-mini
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| o4-miniReasoning Effort=High2025.08 | 87.5 | 0.4 | |
| GPT-5Reasoning Effort=Low2025.08 | 87.4 | — | |
| o4-miniReasoning Effort=Low2025.08 | 87.1 | — | |
| Gemini-2.5-ProReasoning Effort=Low2025.08 | 86.4 | — | |
| GPT-5Reasoning Effort=High2025.08 | 86.1 | -1.3 | |
| Claude-Sonnet-4Reasoning Effort=Low2025.08 | 85.8 | — | |
| Claude-Sonnet-4Reasoning Effort=High2025.08 | 85.8 | 0 | |
| Gemini-2.5-ProReasoning Effort=High2025.08 | 85.1 | -1.3 | |
| o3Reasoning Effort=Low2025.08 | 84.8 | — | |
| o3-miniReasoning Effort=Low2025.08 | 83.8 | — | |
| o3-miniReasoning Effort=High2025.08 | 83.4 | -0.4 | |
| o3Reasoning Effort=High2025.08 | 82.7 | -2.1 |