List your company

Agent Trajectory Verification Agreement on WebTailBench

7.7Unterminated RateNative Verifier
7.3289.83912.3514.861Apr 5, 2026

Evaluation results

MethodLinks
Native VerifierAgent Model=GPT-5, N (tasks scored)=5972026.047.762.563.539.923377278401749646531
Native VerifierAgent Model=Fara-7B, N (tasks scored)=5992026.041739.639.623.230207670501425776449