ByteDance Releases Seedance 2.5: Native 30s Video and Multimodal Control

On July 31, ByteDance's Jimeng (Dreamina) officially released Seedance 2.5, a new video generation model focusing on long-form storytelling and precise multimodal control. It natively generates 30-second videos and offers a long-video mode extendable to 3 minutes. The product is now live on Jimeng and Higgsfield, significantly improving AI video creation workflows.

Confirmed

Unconfirmed

Why it matters

2026-07-30 ~ 2026-08-01 · 60 related posts

Full story(20 episodes)→

Primary sources

22 near-duplicate retellings: Med1_Ai · kimmonismus · mark_k · alifcoder · 字节跳动Seed · koltregaskes · 量子位 · Scobleizer · kimmonismus · mhdfaran · EXM7777 · Eric520CC · thesaraharminta · Div_pradeep · SimplyAnnisa · thetripathi58 · iamfakhrealam · hey_abusiddik · xiaohu · 卡尔的AI沃茨 · nikola_mr64990 · hey_abusiddik