Seedance 2.5
AI video model supporting 30-second single-clip generation and multimodal references
About Seedance 2.5
Seedance 2.5 is the latest model in the Seedance video series and is available on JXP. It supports continuous single-clip video generation up to 30 seconds without stitching; one generation can combine up to 50 text, image, audio, video, and 3D white-model references, with post-generation local replacement of clothing, object repair, or environment reshaping. JXP also provides 480p, 720p, and 1080p output, plus common aspect ratios including 16:9, 9:16, and 1:1.
Best for
Best use cases
Story sequence production
Product demonstrations
Vertical content for TikTok, Reels, or Shorts
Cinematic sequence production
Features
30-second single-clip video
Generate continuous videos up to 30 seconds without stitching.
50 multimodal references
Combine up to 50 text, image, audio, video, and 3D white-model references in one generation.
Local detail editing
Replace clothing, repair objects, or reshape environments without starting over.
Multiple resolutions and aspect ratios
Supports 480p, 720p, and 1080p, plus 16:9, 9:16, and 1:1 aspect ratios.
Pros
- Up to 30-second single-clip generation, no stitching required
- Supports up to 50 multimodal references
- Edit local details without rerendering the whole clip
- Supports 480p, 720p, and 1080p output
Cons
- Confirm watermark removal and commercial usage rights under the current platform and plan terms