ByteDance Launches Dreamina Seedance 2.5: 30-Second Audio-Video Generation

koltregaskes · x · 2026-07-31

ByteDance's Dreamina has introduced Seedance 2.5, a next-generation audio-video joint generation model. The model focuses on 30-second long video generation, with options to extend twice for richer storytelling.

Additionally, it enhances reference material understanding and interactive editing, supporting professional workflows like white-model control and green-screen editing. The company emphasizes built-in safeguards, including invisible watermarking and C2PA, and it is now available to subscribers across various overseas regions.

Related event: ByteDance Releases Seedance 2.5 Video Model(17 posts)→

Original post →

More from Multimodal

Multimodal channel →