ResearchBenchmarksText-to-Video Generation on VMBench motion-oriented prompts 14Text-to-Video Generation on VMBench motion-oriented prompts 14Follow65.02Aesthetic QualityWan2.154.859257.497160.13562.7729Jan 28, 2026Evaluation resultsMethodMethodLinksAesthetic QualityBackground ConsistencyDynamic DegreeImage QualityMotion SmoothnessSubject ConsistencyTemporal FlickeringQuality ScoreWan2.1Model Version=BaseModel Version=Base2026.0165.0291.5733.3362.8597.590.1173.477.27LTDbase_model=Wan2.1base_model=Wan2.12026.0155.2592.5978.166.497.9391.980.3680.85