You can get results as good or better than Midjourney with stable diffusion…but it’s not as streamlined as Midjourney is. you have to find the right model on Civitai.com, you have to craft the prompt right, you can use controlnet. And then if you want to go the extra mile, refine it with img2img.