OJBench (Python)
Code · 2025-06-19
OJBench's 232 NOI and ICPC competition problems solved in Python, judged by hidden test cases under the benchmark's time limits. Scores report pass@1, the percentage of problems whose generated Python program passes every test; Python results run below C++ results on the same problems, so the two are kept separate. Higher is better.
Top models (higher is better)
| Model | Score |
|---|---|
| Gemini 3.1 Pro Preview | 70.7 |
| Kimi K2.6 | 60.6 |
| Opus 4.6 | 60.3 |
| Kimi K2.5 | 54.7 |