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

Code on APT-Bench

41.9Accuracy

Qwen3

22.86827.80932.7537.691Dec 31, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.12
41.9
2025.12
37.9
2025.12
32.8
2025.12
25.1
2025.12
24.3
2025.12
23.6