Action Prediction on AndroidControl Low v2
88.9Pass@1 Step AccuracyGUI-Libra-8B
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| GUI-Libra-8BModel Category=Open-source2026.05 | 88.9 | — | — | 91.7 | — | — | — | — | |
| Mem-W-8BBackbone=UI-Venus-1.5-8B, Memory Augmentation=Mem-W2026.05 | 87.94 | 93.72 | 89.24 | 94.22 | 98.24 | 94.62 | — | — | |
| GPT-5 + UGroundModel Category=Proprietary, Grounding Module=UGround2026.05 | 86.2 | — | — | 90 | — | — | — | — | |
| GUI-Libra-7BModel Category=Open-source2026.05 | 85.2 | — | — | 90.7 | — | — | — | — | |
| Mem-W-4BBackbone=Qwen3-VL-4B, Memory Augmentation=Mem-W2026.05 | 84.92 | 90.95 | 92.58 | 89.2 | 93.72 | 93.72 | — | — | |
| Qwen3-VL-32BModel Category=Open-source2026.05 | 84.42 | 89.95 | 91.83 | 88.69 | 92.21 | 94.17 | — | — | |
| UI-Venus-1.5-8BModel Category=Baseline2026.05 | 81.66 | 88.44 | 91.03 | 87.19 | 93.22 | 94.17 | — | — | |
| UI-Venus-1.5-30B-A3BModel Category=Open-source2026.05 | 79.65 | 87.94 | 87.44 | 87.44 | 91.21 | 91.93 | — | — | |
| Qwen3-VL-4BModel Category=Baseline2026.05 | 78.9 | 85.93 | 90.58 | 82.4 | 88.69 | 93.72 | — | — | |
| GPT-4o + UGroundModel Category=Proprietary, Grounding Module=UGround2026.05 | 78.4 | — | — | 85.4 | — | — | — | — | |
| GPT-4.1 + UGroundModel Category=Proprietary, Grounding Module=UGround2026.05 | 78.4 | — | — | 83.2 | — | — | — | — | |
| Qwen3-VL-2BModel Category=Open-source2026.05 | 77.64 | 85.18 | 87 | 84.92 | 89.95 | 93.72 | — | — | |
| Qwen3-VL-8BModel Category=Open-source2026.05 | 77.6 | 82.91 | 90.27 | 83.2 | 86.93 | 94.17 | — | — | |
| GUI-Owl-1.5-8BModel Category=Open-source2026.05 | 77.14 | 86.68 | 88.34 | 84.42 | 93.47 | 92.38 | — | — | |
| GPT-5-mini + UGroundModel Category=Proprietary, Grounding Module=UGround2026.05 | 77.1 | — | — | 83.2 | — | — | — | — | |
| Step-GUI-4BModel Category=Open-source2026.05 | 68.84 | 75.13 | 88.79 | 82.16 | 86.68 | 94.62 | — | — | |
| GLM-4.1V-9B-ThinkingModel Category=Open-source2026.05 | 67.1 | 90.2 | 67.26 | 73.4 | 93.22 | 78.48 | — | — | |
| GUI-R1-7BModel Category=Open-source2026.05 | 62.3 | — | — | 72.6 | — | — | — | — | |
| GUI-R1-3BModel Category=Open-source2026.05 | 55.8 | — | — | 71.9 | — | — | — | — | |
| MAI-UI-2BModel Category=Open-source2026.05 | 49.25 | 64.32 | 82.51 | 54.77 | 68.09 | 87 | — | — | |
| UI-TARS-1.5-7BModel Category=Open-source2026.05 | 48.5 | 71.36 | 80.27 | 70.9 | 89.95 | 87.44 | — | — | |
| Aguvis-7BModel Category=Open-source2026.05 | 48 | — | — | 48.7 | — | — | — | — | |
| Aguvis-7BType=Open-source Native2026.02 | — | — | — | — | — | — | 48 | 48.7 | |
| GLM-4.1V-9B-ThinkingType=Open-source Native2026.02 | — | — | — | — | — | — | 67.1 | 73.4 | |
| GPT-4.1 + UGround-v1-7BFramework=SeeAct-V, Type=Proprietary2026.02 | — | — | — | — | — | — | 78.4 | 83.2 | |
| GPT-4o + UGround-v1-7BFramework=SeeAct-V, Type=Proprietary2026.02 | — | — | — | — | — | — | 78.4 | 85.4 | |
| GPT-5 + UGround-v1-7BFramework=SeeAct-V, Type=Proprietary2026.02 | — | — | — | — | — | — | 86.2 | 90 | |
| GPT-5-mini + UGround-v1-7BFramework=SeeAct-V, Type=Proprietary2026.02 | — | — | — | — | — | — | 77.1 | 83.2 | |
| GUI-Libra-3BType=Ours2026.02 | — | — | — | — | — | — | 85.9 | 90.5 | |
| GUI-Libra-4BType=Ours2026.02 | — | — | — | — | — | — | 86.4 | 93 | |
| GUI-Libra-7BType=Ours2026.02 | — | — | — | — | — | — | 85.2 | 90.7 | |
| GUI-Libra-8BType=Ours2026.02 | — | — | — | — | — | — | 88.9 | 91.7 | |
| GUI-R1-3BType=Open-source Native2026.02 | — | — | — | — | — | — | 55.8 | 71.9 | |
| GUI-R1-7BType=Open-source Native2026.02 | — | — | — | — | — | — | 62.3 | 72.6 | |
| Qwen2.5-VL-32BType=Open-source Native2026.02 | — | — | — | — | — | — | 78.4 | 85.4 | |
| Qwen2.5-VL-3BType=Baseline2026.02 | — | — | — | — | — | — | 71.1 | 79.2 | |
| Qwen2.5-VL-72BType=Open-source Native2026.02 | — | — | — | — | — | — | 82.9 | 90.2 | |
| Qwen2.5-VL-7BType=Baseline2026.02 | — | — | — | — | — | — | 67.8 | 81.7 | |
| Qwen3-VL-32BType=Open-source Native2026.02 | — | — | — | — | — | — | 83.9 | 86.9 | |
| Qwen3-VL-4BType=Baseline2026.02 | — | — | — | — | — | — | 78.9 | 82.4 | |
| Qwen3-VL-8BType=Baseline2026.02 | — | — | — | — | — | — | 77.6 | 83.2 | |
| UI-TARS-1.5-7BType=Open-source Native2026.02 | — | — | — | — | — | — | 48.5 | 70.9 |