AI video tools are improving fast. But rolling them out isn’t always straightforward.
ByteDance seems to know that.
The company has started adding its new model, Dreamina Seedance 2.0, to CapCut. It’s not a full release yet—only users in a few countries can try it for now.
The feature is meant to make video creation feel easier and quicker. You don’t need much to get started. A short text prompt, an image, or even an existing clip can be enough. In some cases, just describing a scene in a few words works.
It also goes beyond basic generation. The tool can tweak and improve videos—things like movement, lighting, and overall look. For creators, that could mean less time fixing details and more time experimenting with ideas.
Right now, it only supports short clips, up to 15 seconds. That may feel limited, but it’s enough for testing concepts or creating quick content for social platforms.
The slow rollout doesn’t seem random. There have been ongoing concerns around copyright, especially when it comes to AI-generated content. Some reports suggested the global launch was delayed for this reason, and that might explain why access is restricted for now.
ByteDance says it has put some limits in place. The model won’t generate videos using real people’s faces, and it’s designed to avoid unauthorised use of copyrighted material. There’s also an invisible watermark added to the videos it creates, which helps track where they come from.
Interestingly, the same model is already live in China through Jianying, another ByteDance app. So this isn’t entirely new—it’s more like a gradual expansion.
For now, it feels like ByteDance is testing the waters. The tech is ready, but the company seems careful about how and where it rolls it out.
And given how quickly AI video is evolving, that caution probably isn’t a bad thing.
