Text-to-Code generation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
22.65EM
9
Feb 26, 2026
10.6Manim Score
8
Apr 15, 2026
86Task Accuracy
8
Apr 15, 2026
9.7Code Error Rate
8
Apr 15, 2026
5.71BLEU-4
7
Feb 27, 2026
6.55BLEU-4
7
Feb 27, 2026
8.08BLEU-4
7
Feb 27, 2026
2.51BLEU-4
7
Feb 27, 2026