GUI planning on AndroidControl High
75.8SRSE-GA
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| SE-GAModel Size=7B2026.05 | 75.8 | — | 84.4 | 76.4 | |
| UI-TARSModel Size=72B2026.05 | 74.7 | — | 83.7 | 81.5 | |
| OS-AtlasModel Size=7B2026.05 | 71.2 | — | 85.2 | 78.5 | |
| UI-TARS 1.0* + ICMBase Model=UI-TARS 1.0, Refinement=ICM2026.01 | 67.5 | 79.5 | — | — | |
| UI-TARS 1.0* + ICM-r2Base Model=UI-TARS 1.0, Refinement=ICM-r22026.01 | 67.1 | 78.9 | — | — | |
| InternVL-2Model Size=4B2026.05 | 66.7 | — | 84.1 | 72.7 | |
| Doubao + ICM-r2Base Model=Doubao, Refinement=ICM-r22026.01 | 66 | 75.5 | — | — | |
| UI-TARS 1.5* + ICM-r2Base Model=UI-TARS 1.5, Refinement=ICM-r22026.01 | 65.6 | 74.7 | — | — | |
| UI-TARS 1.5* + ICMBase Model=UI-TARS 1.5, Refinement=ICM2026.01 | 64.5 | 73.3 | — | — | |
| Doubao + ICMBase Model=Doubao, Refinement=ICM2026.01 | 64.4 | 75.1 | — | — | |
| Qwen 2.5 VL + ICM-r2Base Model=Qwen 2.5 VL, Refinement=ICM-r22026.01 | 63.8 | 75.9 | — | — | |
| Qwen 2.5 VL + ICMBase Model=Qwen 2.5 VL, Refinement=ICM2026.01 | 63.4 | 75.9 | — | — | |
| DoubaoMethod Type=N/A, Size=N/A2026.01 | 62.4 | 75 | — | — | |
| AguvisMethod Type=SFT, Size=7B2026.01 | 61.5 | — | — | — | |
| Qwen 2.5 VLMethod Type=N/A, Size=7B2026.01 | 60.6 | 70.9 | — | — | |
| SeeClickMethod Type=SFT, Size=9.6B2026.01 | 59.1 | 62.9 | — | — | |
| SeeClickModel Size=9.6B2026.05 | 59.1 | — | 82.9 | 62.9 | |
| UI-TARS 1.0*Method Type=N/A, Size=7B2026.01 | 58.2 | 68.9 | — | — | |
| UI-TARS 1.5*Method Type=N/A, Size=7B2026.01 | 55.8 | 68.1 | — | — | |
| Qwen2.5-VLModel Size=7B2026.05 | 47 | — | 68.6 | 59.7 | |
| GUI-R1Method Type=RFT, Size=3B2026.01 | 46.6 | 56.2 | — | — | |
| UI-R1Method Type=RFT, Size=3B2026.01 | 45.4 | 55.7 | — | — | |
| OS-Atlas-BaseMethod Type=ZS, Size=7B2026.01 | 29.8 | 54.9 | — | — | |
| ClaudeModel Size=-2026.05 | 20.8 | — | 63.7 | 0 | |
| GPT4o + ICM-r2Base Model=GPT4o, Refinement=ICM-r22026.01 | 18.8 | 6.1 | — | — | |
| GPT4o + ICMBase Model=GPT4o, Refinement=ICM2026.01 | 17 | 5.3 | — | — | |
| GPT4oMethod Type=N/A, Size=N/A2026.01 | 13.2 | 3.6 | — | — | |
| GPT-4oModel Size=-2026.05 | 12.5 | — | 66.3 | 0 | |
| Aria-UIMethod Type=SFT, Size=3.9B2026.01 | 10.2 | 43.2 | — | — | |
| Aria-UIModel Size=23B2026.05 | 10.2 | — | — | 43.2 |