List your company

Agent Trajectory Verification Agreement on WebVoyager

3.4Unterminated Rate (%)Native Verifier
3.3683.5843.84.016Apr 5, 2026

Evaluation results

MethodLinks
Native VerifierAgent Model=GPT-5, N (tasks scored)=5932026.043.490.679.47146883900.3627278860.33
Native VerifierAgent Model=Fara-7B, N (tasks scored)=5942026.044.274.64937.965669750.3816063680.33