While OpenAI may be scaling back its efforts in the video generation market with the closure of its Sora app, ByteDance on Thursday confirmed that its new audio and video model, Dreamina Seedance 2.0, will be deployed on its editing platform CapCut.
According to ByteDance, this model allows creators to draft, edit, and sync video and audio content using prompts, images, or reference videos.
The phased rollout will begin with CapCut users in Brazil, Indonesia, Malaysia, Mexico, Philippines, Thailand, and Vietnam, with more markets added over time.
News of CapCut’s launch follows recent reports that global rollout of the model will be paused while the company attempts to address intellectual property issues that have drawn criticism from Hollywood over alleged copyright infringement. This may explain the limited number of markets in which this model is currently available within CapCut.
In China, this model is available to users of ByteDance’s Jianying app.
ByteDance said in the announcement that it works without reference images, even if the creator only uses a few words to describe the scene in their head. CapCut also excels at rendering realistic textures, movement, and lighting across a variety of visual perspectives and angles, and can be used to edit, enhance, and modify creators’ own footage, the company says.
Another use case is to allow creators to test potential ideas based on early concepts and sketches before shooting the actual video.
Additionally, the company explains that Dreamina Seedance 2.0 can be used for a wide range of content where AI video models have historically faced challenges, including videos with cooking recipes, fitness tutorials, business or product overviews, and content focused on motion and action.
At launch, this model supports clips up to 15 seconds in length in six aspect ratios.
CapCut deploys models across different areas, including editing features such as AI video and generation tools such as Video Studio. It will also be introduced to ByteDance’s AI generation platform “Dreamina” and marketing platform “Pippit.”
ByteDance says this model does not have the ability to create videos from images or videos containing real faces, as they have added safety restrictions considering its ability to create realistic content. CapCut also blocks the use of unauthorized generation of intellectual property. (However, with the restrictions working properly, this model is currently available in the US. Perhaps further adjustments could be made.)
ByteDance added that content created with Dreamina Seedance 2.0 will also include an invisible watermark, which will help identify content created with that model when shared outside the platform. This could be useful, such as for takedown requests from rights holders, if the model allowed copyrighted content to pass through.
ByteDance says it will partner with experts and the creative community to iterate and improve the model’s features as it rolls out.
