No clear benefit of using this product. Might as well use Dalle2/Midjourney/Stable Diffusion directly.
Edit: You guys can improve my training a version of SD specifically trained for creating logos. I believe the current model you are using is very less restricted which results in logo not being up to the mark.
