Seedance 2.5 Released: Supports 50 Reference Inputs and 30-Second Videos
SucceededMind · x · 2026-08-07
The AI video generation model Seedance 2.5 has officially launched on the Lovart platform. This update introduces several core feature upgrades, significantly improving workflow efficiency and visual consistency for creators.
Key Updates:
- Multimodal Reference Inputs: Supports up to 50 reference assets (characters, products, styles) to ensure visual consistency across shots.
- Longer Video Output: Capable of generating videos up to 30 seconds long, handling multi-shot scenes, multiple characters, and continuous storytelling.
- Advanced Video Editing: Features timeline-based local editing, allowing users to tweak lighting, mood, or modify specific segments without regenerating the entire video.
- Cinema-Grade Quality: Natively supports high-definition imagery and cinematic visual output.
Additionally, Lovart's built-in agent workflow allows users to generate matching assets like social media cutdowns, thumbnails, and product shots directly within the same project workspace.
Related event: Seedance 2.5 Launches on Lovart with 50 Reference Images Support(3 posts)→
More from Multimodal
- Testing Seedance 2.5: 'Claude on a Japanese TV Show' Yields Impressive Results — mark_k · 2026-08-07
- ByteDance's Seedance 2.5 Enterprise API Goes Live with 30-Second Video Generation — hey_abusiddik · 2026-08-07
- Seedance 2.5 Tested: Generating Coherent Martial Arts and Complex Interactions — ZabihullahAtal · 2026-08-07
- ByteDance's Seedance 2.5 lands on Scenario with 50 multimodal references support — aziz4ai · 2026-08-07
- Mediabunny optimizes large file streaming with on-demand downloads and memory control — Vjeux · 2026-08-07
- SEEDANCE 2.5 Tested: Generating Anime-Style AI Video — LudovicCreator · 2026-08-07