ResearchTasksEnterprise interface task completionEnterprise interface task completionFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedWorkArena L1Gemini 3 Pro79.7Task Success Rate14Apr 10, 2026WorkArena++ L2Gemini 3 Pro41.6Success Rate9Apr 10, 2026