MV-Forcing introduces a unified diffusion framework that combines temporal and view‑wise autoregression to generate long, multi‑view consistent videos of dynamic scenes. By embedding a 4D geometric bridge between sequentially generated views, the method addresses the current gap between long single‑view generation and short multi‑view synthesis. The approach enables coherent spatio‑temporal self‑forcing across both time and camera angles, offering a scalable solution for complex multi‑view video synthesis.