Stable Diffusion is the most flexible AI image generator currently available and it is also entirely open source.
As a result it's also being used as the basis for other 3rd party image generators. Recent updates to the model are producing images that getting closer and closer to the quality of output from Midjourney or Dalle-3. You can try out the new SDXL model that's creating stunning imagery via Stability AI's Dreamscope site.
You can train your own models based on your own dataset to get it to generate exactly the kind of images you're after. Plus unlike the other options here you can run it on your very own PC/Mac and generate as many images as you want, helping to bring generation costs down.