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

Mathematical Reasoning on GSM8K (Avg@5)

97.1Avg@5 Success Rate

GPT-5-high

86.1889.01591.8594.685Jan 23, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.01
97.1
2026.01
96.9
2026.01
96.9
2026.01
96.3
2026.01
96.2
2026.01
96
2026.01
95.9
2026.01
95.8
2026.01
95.6
2026.01
95.6
2026.01
95.6
2026.01
95.6
2026.01
95.5
2026.01
95.4
2026.01
95.3
2026.01
95.1
2026.01
95.1
2026.01
95
2026.01
95
2026.01
95
2026.01
95
2026.01
94.9
2026.01
94.8
2026.01
94.8
2026.01
94.7
2026.01
94.5
2026.01
94.5
2026.01
94.3
2026.01
94.1
2026.01
94
2026.01
93.9
2026.01
86.6