List your company

Trajectory completion judgment on OGRBench MacOS

93.5AccuracyOS-Themis
70.51676.48382.4588.417Jun 5, 2026

Evaluation results

MethodLinks
OS-ThemisVerifier=Gemini-3-Flash2026.0693.583.9
StainFlowVerifier=Gemini-3-Flash2026.0693.584.9
ZeroGUIVerifier=Qwen3.5-27B2026.0692.284.2
ZeroGUIVerifier=Qwen3.5-9B2026.0690.982.1
OS-ThemisVerifier=Qwen3.5-9B2026.0690.977.4
StainFlowVerifier=Qwen3.5-9B2026.0690.980
StainFlowVerifier=Qwen3.5-27B2026.0690.980
ZeroGUIVerifier=GPT-52026.0689.679
OS-ThemisVerifier=Qwen3.5-27B2026.0689.673.3
OS-ThemisVerifier=GPT-52026.0689.673.3
StainFlowVerifier=GPT-52026.0689.680
DigiRLVerifier=Gemini-3-Flash2026.0685.766.7
ZeroGUIVerifier=Gemini-3-Flash2026.0685.771.8
DigiRLVerifier=GPT-52026.0683.131.6
OS-ThemisVerifier=Qwen3-VL-8B2026.0683.151.9
DigiRLVerifier=Qwen3.5-9B2026.0681.822.2
DigiRLVerifier=Qwen3-VL-8B2026.0680.511.8
DigiRLVerifier=Qwen3.5-27B2026.0680.511.8
StainFlowVerifier=Qwen3-VL-8B2026.0680.565.1
ZeroGUIVerifier=Qwen3-VL-8B2026.0671.457.7