ResearchTasksSpoken DialogueSpoken DialogueFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedURO-Bench Chinese Basic TrackStep-Audio 2 mini99.83Repeat Score15Feb 26, 2026URO-Bench Basic TrackGLM-4-Voice85.82Understanding Accuracy12Mar 26, 2026MultiDialog 1.0 (test)SpeechGPT930.401PPL8Feb 26, 2026Big Bench Audio (test)MiMo-Audio-7B-Instruct72.9S2T Accuracy6Feb 26, 2026MultiChallenge Audio (test)MiMo-Audio-7B-Instruct15.15S2T Score5Feb 26, 2026