Code Generation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
88.41HumanEval+ Accuracy
20
Jun 23, 2026
52.19Accuracy
20
Jun 11, 2026
66.7Accuracy
20
Jun 2, 2026
48.96CodeX Score
20
Jun 1, 2026
5.89Speedup
20
May 29, 2026
62.9Pass@4
20
Jul 3, 2026
46.7Mean Accuracy
20
May 26, 2026
83.7Mean Accuracy
20
May 26, 2026
76.8Mean Accuracy
20
May 26, 2026
0.597NLL
20
May 22, 2026
72.1Pass@1
20
Jun 5, 2026
7.23Avg Acceptance Length (τ)
20
May 15, 2026
42.4Test Accuracy (%)
20
May 13, 2026
21.1Avg@4
20
May 12, 2026
3.92Speedup
20
May 11, 2026
49.5pass@1
20
Jun 9, 2026
86.95Pass@1
20
Feb 26, 2026
91Pass@1
20
Feb 26, 2026
56.8Accuracy
20
Feb 26, 2026
67.1Pass@1 (C++)
20
Feb 26, 2026
38pass@1
20
Mar 4, 2026
1.91Speedup vs AR
19
May 27, 2026
32.8LCBv6 Score
19
May 13, 2026
61.7Pass@1
19
Apr 16, 2026
95.9Easy Score
19
Mar 10, 2026