Code Generation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
4.7Speedup (x)
8
Feb 26, 2026
38.2Accuracy
8
Feb 26, 2026
46.1Pass@1
8
Feb 26, 2026
38.12CodeBLEU
8
Feb 26, 2026
30.29Pass@1
8
Feb 26, 2026
69.62Pass@1
7
Jul 7, 2026
55MBPP Score
7
Jul 2, 2026
63.41HumanEval Score
7
Jul 2, 2026
90.85HumanEval+ Accuracy
7
Jun 15, 2026
58.86LCB.v6 Score
7
Jun 2, 2026
1Spearman rho Agreement (FullPair)
7
May 29, 2026
43Pass@1
7
May 26, 2026
61Pass@1
7
May 26, 2026
64pass@1
7
May 26, 2026
65.1Pass@1
7
May 19, 2026
61.92Mean@8
7
May 14, 2026
88.7MBPP+ Accuracy
7
May 13, 2026
24.2BigCodeBench Score
7
May 27, 2026
81AUROC
7
May 1, 2026
48Occ
7
May 1, 2026
74.8pass@1 Accuracy
7
Apr 2, 2026
61.2Pass@1
7
Mar 4, 2026
0.12Training Cost (USD)
7
Feb 26, 2026
0.009Cost ($)
7
Feb 26, 2026
86.9ASR
7
Feb 26, 2026