ResearchBenchmarksVideo Generation on VLM Evaluation SuiteVideo Generation on VLM Evaluation SuiteFollow8.25Aesthetic AppealI-Max5.135.946.757.56Mar 24, 2026Evaluation resultsMethodMethodLinksAesthetic AppealDetail RichnessText AlignmentOverall ScoreI-Max2026.038.257.759.258.42HiFLow2026.03889.258.42Ours2026.0388.259.258.5Real-ESRGAN2026.037.7578.57.75UAV2026.037.57.258.57.75CineScale2026.037.56.757.57.25T3-Video2026.036.756.757.256.92Wan2.22026.035.254.55.255