Seedance 2.0

Seedance 2.0, also referred to as Seedance 2, is a video generation model from ByteDance, first observed on February 16, 2026, on the Jimeng AI platform. It supports text-to-video and image-to-video generation. Early user-facing output was observed at 720p resolution and 24 frames per second, but no official technical report, model card, or benchmark results have been published, leaving architecture and training data composition unconfirmed. On February 18, 2026, the Motion Picture Association (MPA) released a public statement specifically naming Seedance 2.0 in a submission to the U.S. Trade Representative, alleging it was trained on copyrighted works without authorization. Seedance 2.0 matters now because its quiet rollout by a major technology company, absent standard transparency disclosures, combined with a formal and dated industry objection, represents a concrete flashpoint in the escalating conflict between generative AI developers and rights holders over training data provenance and copyright liability.

Two-hop subgraph: this entity, every entity it directly relates to, and every entity those neighbors relate to. Drag a node, scroll to zoom, click to inspect — or click any neighbor and re-center the atlas there.