ResearchTasksEnterprise interface interactionFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedWorkArena L2 full benchmarkGPT-569.4Success Rate3Apr 10, 2026WorkArena L2 (test)A3-Qwen3.5-9B9.7Success Rate2Apr 10, 2026