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

Software Engineering on SWE-bench Verified (Accuracy)

62.6Accuracy

JoyAI-LLM Flash

-2.08814.70631.548.294Oct 28, 2025Dec 9, 2025Jan 20, 2026Mar 3, 2026Apr 14, 2026May 26, 2026Jul 8, 2026
Updated 16d ago

Evaluation Results

MethodLinks
2026.04
62.6
2026.03
61.33
2026.04
59.4
2026.04
57.4
2026.02
52.6
2026.02
51.5
2026.02
48.1
2026.02
47.3
2026.02
45.6
2026.02
44.9
2026.02
42.8
2025.10
40.3
2025.10
40.2
2026.06
38.9
2026.06
37.6
2025.10
36.8
2025.10
34.4
2025.10
33.6
2026.06
31.9
2026.04
31.2
2025.10
30.6
2026.07
29.8
2026.04
29
2026.07
27
2026.07
23
2026.06
22.7
2026.06
22.1
2025.10
20.6
2025.10
20.4
2025.10
20.2
2025.10
16.4
2025.10
15.8
2025.10
15.2
2026.06
13.2
2026.03
12.87
2026.06
12.5
2025.10
10.6
2025.10
10.6
2025.10
5.8
2025.10
2.8
2025.10
2.2
2025.10
2
2025.10
0.4