Kling AI Elements Feature For Consistency

Kling AI has released yet another update of their platform, and this time it's 'Elements'. Currently in early access but likely available soon for all subscribers.

This new thing allows users to introduce specific elements into the video, for example, an image of themselves.

Several video generators have released a similar feature on their platforms not too long ago. Pika and their scene ingredients, Hailuo AI and their character reference and Pollo AI's consistent character videos (using Vidu's tech I understand).

But get this: Kling's elements support up to 4 images! Max I've seen so far was 3, and Hailuo only supports 1 so far. So this is a big improvement. With feature like this, you can skip having to provide a long precise textual description of your characters, objects or scenes and uinstead just upload pictures of them and reference them in your final prompt, which will now get much shorter, containing only the important parts: who is doing what, or what is happening, how, where, what is the camera doing etc.

This new capability would allow ecommerce website owners to generate short videos of their clothes worn by models. Even though Kling does have Virtual Try-On specifically for that, but that's images only, and it doesn't yet cover things like eyewear:

The Elements feature should cover that, as it does in other AI generators, where I was able to successfully make my model put on my specific sunglasses.

So it will be exceptionally useful for companies to create ads for their products of any kind: vehicles, accessories, food and drink - you name it. Product consistency is a huge deal for businesses.