Algorithmic Correctness Evaluation on OJBench
71.5Overall ScoreKeye-VL-2.0-30B-A3B
Evaluation results
| Method | Links | |
|---|---|---|
| Keye-VL-2.0-30B-A3BModel Size=30B-A3B2026.06 | 71.5 | |
| Qwen3.5Model Size=35B-A3B2026.06 | 70.2 | |
| GPT-5-miniRelease Date/Version=2025-08-072026.06 | 58.7 |