Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Shell-based agent task completion on Terminal-Bench V2

86.5Accuracy

LLM-as-a-Verifier

79.94881.64983.3585.051Jul 6, 2026
Updated 19d ago

Evaluation Results

MethodLinks
2026.07
86.5--
2026.07
84.7--
2026.07
80.2--
2026.07
80.2--
2026.07
-83.192.1