A full-fidelity text-to-video model built for cinematic results. Generates multi-shot, 1080p videos with smooth motion, strong prompt adherence, and scene continuity.