Seedance 2.5
Multimodal reference input and 4–30-second audio and video generation
About Seedance 2.5
Seedance 2.5 is an audio and video generation model in the Seedance series. It can generate complete videos up to 30 seconds long and supports up to 50 text, image, audio, video, and textureless 3D model references in one generation. Creators and teams can start with a prompt, add references when they need more control, and iterate on pacing and style. It suits ads, short-form social content, product explainers, training materials, and film previsualization. Before generation, choose 4–30 seconds, an aspect ratio, 480p/720p/1080p, and audio on or off, then preview credit usage.
Best for
Best use cases
Ad production
Short-form social video creation
Product explainers and training materials
Film previsualization
Features
Multimodal references
Add up to 50 text, image, audio, video, and textureless 3D model references to one generation.
Duration and output controls
Choose 4–30 seconds, an aspect ratio, 480p/720p/1080p, and audio on or off.
Prompt iteration workflow
Start with a prompt, add references as needed, and iterate on pacing and style.
Credit preview
Preview the credit usage for the generation before submission.
Pros
- Use up to 50 multimodal references at once
- Generate videos up to 30 seconds long
- Adjust duration, aspect ratio, resolution, and audio
- Preview credit usage before submission
Cons
- Mini Plan does not include the fast generation queue
- Mini Plan does not include a commercial-use license
Pricing Plans
Mini Plan
- For hobbyists and beginners
- For light creation and personal use
- An entry plan for steady output
Standard Plan
- For most creators
- For ongoing publishing, frequent creation, and marketing assets
* Pricing information is for reference only. Please refer to the latest announcement on Seedance 2.5's official website.