| Paper | Code | Element Accuracy | Step Success Rate | ModelName | ReleaseDate |
|---|---|---|---|---|---|
| UI-TARS: Pioneering Automated GUI Interaction with Native Agents | ✓ Link | 74.7 | 68.6 | UI-TARS-72B-SFT | 2025-01-21 |
| Aguvis: Unified Pure Vision Agents for Autonomous GUI Interaction | ✓ Link | 69.5 | 64.0 | AGUVIS-72B | 2024-12-05 |
| STEVE: A Step Verification Pipeline for Computer-use Agent Training | ✓ Link | 46.2 | 40.0 | STEVE UI-Grounding + GPT-4o (extra data) | 2025-03-16 |
| GUICourse: From General Vision Language Models to Versatile GUI Agents | ✓ Link | 36.9 | 33.2 | MiniCPM-GUI + Larger LM (extra data) | 2024-06-17 |
| SeeClick: Harnessing GUI Grounding for Advanced Visual GUI Agents | ✓ Link | 28.3 | 25.5 | SeeClick | 2024-01-17 |
| CogAgent: A Visual Language Model for GUI Agents | ✓ Link | 62.3 | CogAgent-18B | 2023-12-14 | |
| VisCritic: Visual State Comparison as Process Reward for GUI Agents | 54.1 | Qwen2.5-VL Agent + VisCritic | 2026-06-23 |