Not sure why they just not replicate the workflow that nano banana pro uses. It lets the thinking model generate a detailed description and then renders that im

Not sure why they just not replicate the workflow that nano banana pro uses. It lets the thinking model generate a detailed description and then renders that image. When I use ChatGPT thinking model and render an image I also get pretty good results. It's not as creative or flexible as nano banana pro, but it produces really useful results.