Code Generation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
83.5Pass@1
62
Jul 1, 2026
287.77Tokens/s
61
Mar 6, 2026
83Performance (%)
60
May 28, 2026
94.2Pass@1 Accuracy
59
Mar 9, 2026
80.4Pass@1
58
May 27, 2026
59.6Accuracy
57
Jun 9, 2026
84.6Pass@1 Accuracy
55
Jun 18, 2026
87.8Pass Rate
55
May 28, 2026
57.1Accuracy
54
Jun 8, 2026
86.19LiveCode BenchV6
52
May 20, 2026
76.5Average Score
52
Feb 26, 2026
76.42Pass@1
51
Jun 4, 2026
1,784Pass@1
51
Mar 9, 2026
81.7Accuracy
51
Feb 26, 2026
85.7Accuracy
51
Feb 26, 2026
4,290TPS
50
May 29, 2026
58.3pass@1
50
Jun 30, 2026
93HumanEval Score
50
Feb 26, 2026
96.3HumanEval Accuracy
49
May 26, 2026
75.4Pass@1
48
Jun 18, 2026
50.68Pass@1
48
Jun 9, 2026
90.5Top-1 Acc.
48
Jun 12, 2026
28.4Pass@1
48
Mar 16, 2026
0.504Pass@1
48
Mar 16, 2026
72.3Pass@1
48
May 27, 2026