List your company

GUI planning semantic consistency on GUIGuard-Bench PC 1.0

328PC Avg.GPT-5.2
85.68148.59211.5274.41Jan 26, 2026

Evaluation results

MethodLinks
GPT-5.2Category=Closed-source VLM2026.01328
Gemini-3-ProCategory=Closed-source VLM2026.01324
Qwen3-VL-235B-A22BCategory=Open-source VLM2026.01252
GUI-Owl-7BCategory=GUI Agent2026.01247
Claude-Sonnet-4.5Category=Closed-source VLM2026.01242
Fara-7BCategory=GUI Agent2026.01169
UI-Tars-1.5-7BCategory=GUI Agent2026.0195