<aside> 🔗 Stable Diffusion Online Provider (AUTOMATIC1111 Web-UI) (Affiliate link)
</aside>
<aside> 🔗 Stable Diffusion Google Colab Notebook (AUTOMATIC1111 Web-UI)
</aside>
<aside> 🔗 Windows installation guide
</aside>
<aside> 🔗 Mac installation guide
</aside>
Use our site’s generator or a free site if you want to play around and generate a few images.
Models available in our site’s generator.
<aside> 📄 Free sites Many of these sites have paid options for extra functionalities. This option is suitable for users who are looking for an image generation service so that they don’t need to install any software.
Pros ✅ No setup.
✅ Easy to use.
Cons
❎ Limited functionality.
❎ Cost more than other options. (Unless you are using our site’s FREE generator)
This is for you if
👉 You don’t want to deal with troubleshooting software. 👉 You are ok with limited functionality. 👉 You don’t mind paying a bit more for some features.
</aside>
You need to use AUTOMATIC1111 (A1111) GUI if you want to follow my tutorials and get the most out of Stable Diffusion. This is suitable for advanced users or who aspire to be one. There are 4 options for using A1111
<aside> 📄 Online provider Think Diffusion manages AUTOMATIC1111 for you for a fee.
🔗 Start using AUTOMATIC1111 online (Affiliate link)
Pros
✅ You don’t need to maintain your A1111 (it will have problems from time to time)
✅ All models and extensions are immediately available to use.
✅ Don’t use up your computer resources (models can be big)
✅ Cost effective if you account for hardware cost.
✅ Faster startup time than Colab.
Cons
❎ Cost more than Google Colab.
This is for you if
👉 You don’t mind paying a bit more to save yourself from troubles of maintaining the software. 👉 You want the most advanced features. 👉 You don’t have the proper hardware or don’t want to use it.
</aside>
<aside> 📄 Google Colab Colab is Google’s interactive computing platform. We provide a notebook for you to run the Web-UI. You will need to subscribe to Colab Pro to use the Web-UI (~$10/month)
📜 Setup guide.
🔗 Open Google Colab Notebook
Pros
✅ You don’t need to maintain your A1111 (it will have problems from time to time)
✅ Don’t use up your computer resources (models can be big)
✅ Cost effective if you account for hardware cost.
Cons
❎ Slower startup time.
❎ Some extensions may not work.
❎ Breaks from time to time. Will be at Andrew’s mercy to fix it.
This is for you if
👉 You want the most advanced features. 👉 You don’t have the proper hardware or don’t want to use it. 👉 You are tech-savvy enough to deal with model files. 👉 You don’t mind paying a modest fee to use regularly.
</aside>
<aside> 📄 Windows PC You need a discrete NVIDIA graphic card with at least 4GB VRAM.
See the installation guide.
Pros
✅ Free
Cons
❎ Can take up a lot of your disk space.
❎ Not cheap if you decide to buy a good GPU card for it.
This is for you if
👉 You want the most advanced features. 👉 You have the proper hardware on your PC and don’t mind using it for SD. 👉 You are tech-savvy enough to install and maintain software on your PC. 👉 You want a completely free solution.
</aside>
<aside> 📄 Mac You need Apple Silicon M1 or M2.
See the installation guide.
Pros
✅ Free
Cons
❎ Slower than PC
❎ Some features are not available
This is for you if
👉 You want the most advanced features but ok with some features being not available. 👉 You have the right hardware on your PC and don’t mind using it for SD. 👉 You are tech-savvy enough to install and maintain software on your PC. 👉 You want a completely free solution. 👉 You don’t mind image generation is a bit slow on Mac.
</aside>
Text-to-image is the most basic usage of Stable Diffusion. It is a model trained to turn text descriptions into images. For example, a text prompt, “a photo of an astronaut riding a horse on the moon” will produce an image just like that.
However, it could be challenging to generate precisely the image you want. We need to learn particular keywords it was trained on to entice it to give you what you want.
💡 TIP 1: Be as detailed and as specific when describing a subject
<aside> ❌ A woman sitting
</aside>
<aside> ✔️ A beautiful woman with blue eyes and blonde hair sitting outside in a restaurant
</aside>
💡 TIP 2: Use names of artists and websites to modify the style
<aside> ❌ digital art
</aside>
<aside> ✔️ digital art, artstation, artgerm, alphonse mucha
</aside>
💡 TIP 3: Add lighting terms to make it more interesting
Extra Tip: Always generate a few images with the same prompt to understand what the prompt can do!
<aside> ⭐ Check out Stable Diffusion prompt builder for a complete step-by-step system.
</aside>