| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| OSWorld 1.0 (test) | Success Rate (OS)91.67 | 42 | 25d ago | ||
| DataScope V3 | GPT-5.2+ UIIns | Success Rate (SR)3.1 | 10 | 25d ago | |
| DataScope V2 | GPT-5.2+ UIIns | Success Rate (SR)6.2 | 10 | 25d ago | |
| DataScope V1 | GPT-5.2+ UIIns | Success Rate (SR)18.8 | 10 | 25d ago | |
| WeChat Mini-App Synthetic (train val) | code-native-reward (Ours) | VLM Eval Success Rate71.81 | 4 | 5mo ago | |
| WeChat Mini-App Real-World (test) | code-native-reward (Ours) | Success Rate (Human)42.28 | 4 | 5mo ago |