List your company

GUI planning semantic consistency on GUIGuard-Bench Overall 1.0 (Mixed)

1.16Black Mask ScoreUI-Tars-1.5-7B
1.07321.65912.2452.8309Jan 26, 2026

Evaluation results

MethodLinks
UI-Tars-1.5-7BCategory=GUI Agent2026.011.161.440.881.011.11
Fara-7BCategory=GUI Agent2026.012.021.942.111.952.04
GUI-Owl-7BCategory=GUI Agent2026.012.562.52.852.322.55
Qwen3-VL-235B-A22BCategory=Open-source VLM2026.012.712.922.632.632.73
Claude-Sonnet-4.5Category=Closed-source VLM2026.012.872.7433.032.96
Gemini-3-ProCategory=Closed-source VLM2026.013.323.423.43.223.35
GPT-5.2Category=Closed-source VLM2026.013.333.33.443.173.31