How do you use a model?
1 Answer
Hi there!
I'm working on a beginners guide right now on my Patreon, but the short answer is that you'll need a moderately powerful PC with ideally an Nvidia graphics card. You can then download and run some software which allows you to easily send text to a model (in simple terms! It's extremely complex behind the scenes!) and get an image back!
You can even do this without an Nvidia card - if you have a Mac, or an AMD, or even just using your computer's CPU (which will be slow).
There are a number of interfaces available, varying in complexity. You have the easiest - DiffusionBee (Mac only) and EasyDiffusion, then some more advanced ones like InvokeAI, or Automatic 1111's WebUI (the most popular), then the super advanced UIs like ComfyUI.
Alternatively, there are web based services like Prodia, Mage.space, and NightCafe (and many others). These services typically host a variety of models, and all the processing happens at their end - you type a prompt and get the image, but they're not as customizable as running Stable Diffusion on your own PC. There also might be some cost involved - you're paying for a service.
I hope this helps! I will get my beginners guide posted as soon as possible!