Seedance 2 is an advanced text-to-video and image-to-video platform that converts natural language prompts and reference media into multi-shot, coherent videos. It accepts multimodal input (up to 9 images and 3 clips), features native audio sync with lip matching, and supports multi-shot narrative composition for consistent characters and scenes. Seedance 2 focuses on speed and quality, delivering 1080p/2K outputs in 30–90 seconds and providing an API and commercial licensing for production and developer integration.