ResearchDatasetsVideoThinkBenchVideoThinkBenchFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsText-centric ReasoningVideoThinkBench mini (test)89Average Score22Vision-centric tasksVideoThinkBench mini (test)37.3Average Score18Video generation reasoningVideoThinkBench (test)44.7Eyeballing-Point Accuracy4