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

Task Goal Completion on AppWorld Hard 63 tasks (test_normal)

31.7Accuracy

No compression

3.72410.98718.2525.513Jun 18, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
31.711.950
2026.06
17.56.580
2026.06
14.37.790
2026.06
4.85.687.37