On July 24th, Keling AI Video unveiled a groundbreaking upgrade to its multi-image reference model and capabilities. Rigorous blind tests have confirmed that the revamped model's effectiveness has soared by an impressive 102%. This latest enhancement ensures that videos crafted with the multi-image reference feature now exhibit heightened character consistency, markedly improved dynamic quality, and more fluid, natural interactions among multiple subjects. Moreover, the overall visual aesthetics have attained a new level of sophistication. Additionally, the multi-image reference functionality now incorporates a novel local content reference feature, empowering users to pinpoint specific subjects, faces, or attire within images. This innovation eliminates distractions from irrelevant elements, resulting in even more polished and perfect video outcomes.