ResearchBenchmarksClassification on MTBench WeatherClassification on MTBench WeatherFollow55.7AccuracyTS-Debate17.53227.44137.3547.259Jan 27, 2026Evaluation resultsMethodMethodLinksAccuracyTS-Debate2026.0155.7Zero-ShotMultimodal Context=trueMultimodal Context=true2026.0152.3CoTMultimodal Context=trueMultimodal Context=true2026.0147.7MADMultimodal Context=trueMultimodal Context=true2026.0132Zero-ShotMultimodal Context=falseMultimodal Context=false2026.0125CoTMultimodal Context=falseMultimodal Context=false2026.0122VL-Time2026.0122MADMultimodal Context=falseMultimodal Context=false2026.0120ByMyEyes2026.0119