ByteDance Launches Seedance 2.5: 30s Generation, 50 Multimodal Reference Assets
SucceededMind · x · 2026-08-04
ByteDance's AI video tool Dreamina officially launched the Seedance 2.5 model, now available to US users at the lowest price point in the Seedance family.
The upgrade significantly enhances creator control and narrative coherence: a single generation can last up to 30 seconds, reducing jarring cuts. It also supports up to 50 multimodal reference assets (images, videos, audio) in one generation. New features like precise partial editing, 3D white model, and green screen support make the workflow closely resemble a professional studio production.
More from Multimodal
- Qwen Image Edit crashes in ComfyUI when blur nodes are added first — UnorthodoxyMedia · 2026-08-04
- MiniMax H3 opens its weights as a multimodal video model with native stereo audio — mark_k · 2026-08-04
- MiniMax H3 R2V follows a single-image 6×6 storyboard well — smereces · 2026-08-04
- MiniMax opens H3 on Hugging Face as a 33B multimodal video model with stereo audio — mark_k · 2026-08-04
- MiniMax H3 turns “CHICKEN GAME” into a deadpan AI joke — the_bollo · 2026-08-04
- WaiT: Wavelet-Aware Flow Matching Sets New SOTA in Pixel-Space FID — chaumian · 2026-08-04