List your company

End-to-End Environment Interaction on OSWorld-Verified (test)

61.4Pass@1Claude-4.5-sonnet
21.46431.83242.252.568Dec 17, 2025

Evaluation results

MethodLinks
Claude-4.5-sonnet2025.1261.4—
UI-TARS-22025.1247.5—
Claude-4-sonnet2025.1243.9—
UI-TARS-1.52025.1242.5—
Step-GUI-8BEval Protocol=Pass@12025.1240.2—
Mobile-Agent-v32025.1237.7—
GLM-4.6V2025.1237.2—
GUI-OwlSize=7B2025.1234.9—
SeedVL-1.52025.1234.1—
Qwen3-VLSize=8B, Variant=Instruct2025.1233.9—
Step-GUI-4BEval Protocol=Pass@12025.1231.9—
Qwen3Size=30B-A3B, Variant=Instruct2025.1230.3—
UI-TARS-1.5Size=7B2025.1227.4—
Qwen3-VLSize=4B, Variant=Instruct2025.1226.2—
UI-TARSSize=72B, Variant=DPO2025.1224—
OpenAI CUA o32025.1223—
Step-GUI-4BEval Protocol=Pass@32025.12—40.4
Step-GUI-8BEval Protocol=Pass@32025.12—48.5