I have found an text to video ai wich is even better than Luma Ai, its called Vidu, the thing is, this ai may sometimes give shitty results but of all the Ai's i ever saw, this one is the only one who tries to make inflation content, normally i have to make "magic" or 1 or even use 2 images like in luma, but this Ai can make it with a single prompt, but i rather recommend still use a image btw, and the inflation that makes looks jiggly, maybe is bc i added "Water balloon", so i recomend your prompts to be like this "action movie. a woman, her whole body bloats and blows up like a gigantic and obese human water balloon woman, her slender shape turns into a bloated and rounder one, her belly expands like a balloon being blown up", but you can make your prompts like you want to. Here i tried to make a video of a human Jasper inflating..but on the design part it didnt end up well, but the inflation looks really good
https://bbw-chan.link/.media/3bf7372085df3b821c9be88f640826272e48e6e6e6a65cff4e8b23595badde21.webm
also sometimes triggers a warning, its slighty annoying, but when that happens i normally modify the prompt a little or keep generating again
another example of what this can do