ResearchTasksExpert-Level Human Knowledge ReasoningFor AI assistantsMarkdownContent indexExpert-Level Human Knowledge ReasoningFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedHumanity's Last ExamTongyi DeepResearch38.3Pass@111May 19, 2026