ResearchTasksLong-answer Visual Question AnsweringFor AI assistantsMarkdownContent indexLong-answer Visual Question AnsweringFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedMMMU-Pro-VQwen2.5-VL-3B26.3Accuracy9Apr 10, 2026CT-RATE (test)CT-CHAT47.02BLEU2Feb 26, 2026