ResearchTasksGeneral AI Assistant TaskFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedGAIA (val)agent-203096.23Level 1 Score97May 29, 2026GAIAStep-3.5-Flash84.5Accuracy36Jun 9, 2026GAIA-127OpenThinkerAgent-32B23.6Accuracy14Jun 24, 2026GAIA (test)MASLab-ReAct46.06Overall Accuracy14Jun 30, 2026