ResearchTasksFunction-level Code GenerationFunction-level Code GenerationFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedHumanEval+ augmented (test)Mercury Coder90Pass@165May 11, 2026MBPP+ augmented (test)LLaDA2.0-flash79.6Pass@156Mar 13, 2026