Stable Diffusion is a versatile open-source AI generative model capable of creating stunning images from text prompts, image prompts, or a combination of both. With plugins, it provides more powerful and precise control, enabling users to unleash their creativity in various ways. It is supported by an active community that has developed numerous exceptional models and plugins.
As Midjourney no longer offers free trials for unpaid users, Stable Diffusion serves as an excellent alternative for those looking to explore and experience a taste of AI generative art.
Unlike Midjourney, which allows interaction via Discord, Stable Diffusion requires initial setup and launching. Fear not! I’m here to guide you through the simplest way to get started. In just a few clicks, you’ll be ready to enjoy Stable Diffusion with the popular WebUI Automatic1111 – no GPU required on your computer!
First, sign up for a free Google account. This gives you access to Google Colab, where you can experiment with Stable Diffusion. However, with a free account, you may experience disconnections during peak times or after extended use. For more power and stability, consider upgrading to a pro subscription at $9.99 per month, which offers around 50 hours of standard GPU time.
Once your Google Colab account is set up, you’ll need to find notebooks that can launch Stable Diffusion as desired. In this context, a notebook is a pre-written script that prepares the necessary resources and launches an instance of Stable Diffusion, which you can access through your web browser.
Thankfully, the open community has made this incredibly easy.
Visit this 🦒 Colab notebook list, where you’ll find links to a variety of notebooks that launch fantastic Stable Diffusion models. By clicking the links in the “Info – Token – Model Page” column, you can view example images and get a sense of what each model does before trying it yourself.
After deciding which model to explore, you can choose from lite, stable, or nightly versions. Here’s a quick breakdown of the differences, according to the official explanation:
- lite: Stable WebUI and installed extensions
- stable: ControlNet, stable WebUI, and installed extensions
- nightly: ControlNet, latest WebUI, and daily installed extension updates
For beginners, the lite version is a great starting point. If you want to try ControlNet, opt for the stable version, while the nightly version offers the most up-to-date extensions.
Once you’re in Google Colab, click the run icon in the top left corner of the code area. It may take around 10 minutes for Stable Diffusion to be ready.
After the scripts finish running, click on one of the provided links to access the WebUI of your Stable Diffusion instance.
Now you can start exploring Stable Diffusion! Numerous online tutorials are available for both beginners and advanced users. As I continue to experiment with Stable Diffusion, I’ll share my experiences in future posts.
Remember to stop the running instance once you’ve finished playing; otherwise, it will continue to consume GPU time if you have a pro subscription. To do this, click on the top right drop-down menu and select “Disconnect and delete runtime” to ensure you’re not using unnecessary resources.
You can also check your remaining GPU time in the menu:
With this easy-to-follow guide, you’re now ready to dive into the world of AI generative art with Stable Diffusion. Happy experimenting!
