List your company

Trajectory completion judgment on OGRBench Web

87.4AccuracyOS-Themis
60.04867.14974.2581.351Jun 5, 2026

Evaluation results

MethodLinks
OS-ThemisVerifier=Qwen3.5-27B2026.0687.486.7
StainFlowVerifier=Qwen3.5-27B2026.0687.487.2
StainFlowVerifier=Gemini-3-Flash2026.0686.286.2
ZeroGUIVerifier=Qwen3.5-27B2026.0685.386.7
ZeroGUIVerifier=Gemini-3-Flash2026.0683.184.2
OS-ThemisVerifier=Qwen3-VL-8B2026.0682.683.6
StainFlowVerifier=GPT-52026.0682.682.2
OS-ThemisVerifier=Gemini-3-Flash2026.0682.581.8
ZeroGUIVerifier=GPT-52026.0681.682.1
OS-ThemisVerifier=Qwen3.5-9B2026.0681.680
StainFlowVerifier=Qwen3.5-9B2026.0680.579.6
ZeroGUIVerifier=Qwen3.5-9B2026.0679.581.2
StainFlowVerifier=Qwen3-VL-8B2026.0679.582.5
ZeroGUIVerifier=Qwen3-VL-8B2026.067982.2
OS-ThemisVerifier=GPT-52026.0677.975.3
DigiRLVerifier=Qwen3.5-27B2026.0668.463
DigiRLVerifier=GPT-52026.0666.861.8
DigiRLVerifier=Gemini-3-Flash2026.0666.763.6
DigiRLVerifier=Qwen3-VL-8B2026.0665.360.7
DigiRLVerifier=Qwen3.5-9B2026.0661.147.1