ResearchTasksVisual Question Answering, Information Extraction, and Element LocalizationFor AI assistantsMarkdownContent indexVisual Question Answering, Information Extraction, and Element LocalizationFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedVisual Understanding BenchmarkCRAFT-GUI-32B94Accuracy4Mar 16, 2026