List your company

Terminal-Bench 2.0 execution on Terminal-Bench 2.0 (dev)

75Pass RateClaude Code
57.663262.164166.66571.1659Jun 10, 2026

Evaluation results

MethodLinks
Claude Code2026.067516.67
Arbor2026.0672.2213.89
Codex2026.0663.895.56
Initial2026.0658.33—