Seedance 2.0

ByteDance[video]

Unified multimodal video model with native audio, director-level camera control, and 4-15s generation.

Per Second

$0.30

of media

Released

2026-04-09

Features

[native-audio][4-15s-videos][720p][end-frame-control][director-camera-control]

The Lab

location: Beijing, Chinafounded: 2023backed by: ByteDance Ltd.

ByteDance Seed is the foundation-model research division of ByteDance, established in 2023 and headquartered in Beijing. The team develops the Doubao AI assistant and the core foundation models powering it, spanning language, video generation, and image synthesis. Seed's models are deployed at scale across ByteDance's consumer products including Douyin and TikTok.

more about ByteDance Seed →

Model ID: seedance-2-0