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

Software Engineering on SWE-Bench Pro (public)

64.3Resolve Rate (Pass@1)

Claude Opus 4.7

41.83647.66853.559.332Dec 11, 2025Jan 10, 2026Feb 9, 2026Mar 12, 2026Apr 11, 2026May 11, 2026Jun 11, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
64.3-----------
2026.06
59.1-----------
2025.12
59-----------
2025.12
56-----------
2025.12
54.3-----------
2026.06
54.2-----------
2025.12
52.7-----------
2025.12
52-----------
2026.06
51.9-----------
2026.06
50.9-----------
2026.05
50.256.263.768.432.935.556.545.638.55072.70
2026.05
495159.368.432.938.256.549.138.551.861.40
2026.05
47.64964.86232.934.254.849.143.15047.70
2026.05
46.952.15564.631.835.55050.938.546.461.40
2026.06
46.1-----------
2025.12
45.8-----------
2025.12
45.5-----------
2025.12
43.6-----------
2025.12
42.7-----------