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

Long-horizon Task Solving on PINCHBENCH 23 tasks original

93.3Best Success Rate

Claude-Opus-4.6

43.27656.26369.2582.237Jun 15, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
93.381.6
2026.06
90.579.4
2026.06
89.882.8
2026.06
89.577.4
2026.06
88.869.7
2026.06
87.480.4
2026.06
86.777.5
2026.06
72.858.9
2026.06
71.556.4
2026.06
7155.7
2026.06
67.553.8
2026.06
67.152
2026.06
6650.3
2026.06
64.949.2
2026.06
64.547.9
2026.06
63.963.3
2026.06
49.835.4
2026.06
46.134.8
2026.06
45.231.8