Math Word Problem Solving
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
81.4Accuracy (pass@1)
6
Jun 8, 2026
65.7Accuracy
6
Feb 26, 2026
48.2Accuracy
6
Feb 26, 2026
81.97Accuracy (GSM8K)
5
May 21, 2026
0.964Accuracy
5
Mar 10, 2026
77.18Accuracy
4
Feb 26, 2026
54.05Accuracy
4
Feb 26, 2026
11.41Accuracy
4
Feb 26, 2026
57.39Accuracy
4
Feb 26, 2026
52.5Accuracy
3
Feb 26, 2026
73.7Human Solve Rate (MWP)
1
Apr 14, 2026
68.2Human-written MWP Solve Rate
1
Apr 14, 2026