# audio‑visual synthesis

Seedance 2.0: ByteDance’s AI Turns Prompts into Sounding, Moving Films — and Rewires the Cost of Production

ByteDance’s Jimeng AI has launched Seedance 2.0, a generative video model that synchronises images and sound and can follow complex camera directions. Independent tests by NetEase show striking gains in action consistency and realistic ambience, but also persistent artifacts, transition roughness and high compute costs. The model promises to lower production marginal costs and expand commercial markets while raising IP, deepfake and regulatory challenges.