ResearchTasksStreaming Long-context Question AnsweringStreaming Long-context Question AnsweringFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedSTEM-Bench The OfficeProStream (-w/o Mpend)0.304Time14Mar 6, 2026STEM-Bench FriendsProStream (-w/o Mtree)0.318Time14Mar 6, 2026STEM-Bench The Big Bang TheoryProStream (-w/o Mtree)0.4Time (s)14Mar 6, 2026