As we all know the world of Stable Diffusion is a treasure trove of artistic potential. But with so many models available, choosing the right one for your project can be daunting. SD3 is now out for testing through API. I am into line art with strong contrasts and vibrant colors and also skulls ewtc are my favs.
Here is the lineup. As even a prompt like "woman wearing a tight dress" was flagged by SD3 API its not possible to evaluate its pros or cons generating bodies/halfbodies etc.
SD3 is very fast in generating images 50% faster then SDXL0.9 with refiner but fe for skull art SDXL0.9 with refiner is top notch for me.
Update 19.04.2024: For text in the image SD3 is superior to all other models I have tested, "stylish text prompts" are leading to extraordinary outputs.