In the video titled “New #1 AI Video Model is HERE (Beats VEO 3)” by Matthew Berman, published on June 28, 2025, the host discusses the latest advancements in AI video generation with a focus on the new Seed Dance 1.0 model by ByteDance. This model reportedly surpasses Google’s VEO 3 in performance, particularly in generating realistic animations and video content. Berman showcases various video examples, comparing Seed Dance and VEO 3 across different scenarios including street interviews, action sequences, and animation styles. He emphasizes the importance of prompt structure for optimal results and highlights the potential of these tools for content creators. The video concludes with an invitation to try out the new model and a discussion of the implications for AI in video production.