5-10s cinematic video, text or image driven.
Seedance is Volcengine's Doubao video model. Async task-based β 45-120s typical latency. Supports both text-to-video and image-to-video with strong motion coherence and prompt adherence.