Stability AI has publicly released its open source Artificial Intelligence text-to-image model called Stable Diffusion, which is accessible through various platforms with different user interfaces.
Abstract
Stability AI has announced the public release of its open source Artificial Intelligence text-to-image model called Stable Diffusion. The model, which is available for free with weights, can be used with local GPUs. The public can access a demo tool in a Hugging Faces space, but it can be slow due to the number of users. A Google Colab tool is available for those who don't want to wait, but it has a less user-friendly UI compared to Hugging Faces. Dream Studio is a paid version of Stable Diffusion that offers more control over parameters and can be accessed with a monthly subscription. The model is released under a Creative ML OpenRAIL-M license, which allows for commercial and non-commercial usage.
Bullet points
Stability AI has released its open source Artificial Intelligence text-to-image model, Stable Diffusion
The model is free and can be used with local GPUs
The public can access a demo tool in a Hugging Faces space
A Google Colab tool is available for a faster experience, but has a less user-friendly UI
Dream Studio is a paid version of Stable Diffusion with more control over parameters and can be accessed with a monthly subscription
The model is released under a Creative ML OpenRAIL-M license, allowing for commercial and non-commercial usage.
Stable Diffusion Finally Released To The Public
Image from Stability AI
Stability AI has announced the public release of its open source Artificial Intelligence text-to-image model called Stable Diffusion.
This is huge.
Yesterday, the CEO and founder of the company, Emad Mostaque, made a public announcement on Twitter.
“Proud of everyone involved in releasing this tech that is the first of a series of models to activate the creative potential of humanity,” he said
Why should you care?
It’s open source.
It means you can get it from the internet for free — weights and all.
You can use your local GPU to run it. This means that whether you like it or not, the latest trends in text-to-art AI models are both more available and more transparent.
Today we all made the World a more creative, happier and communicative place. More to come in the next few days but I for one can’t wait to see what you all create. Let’s activate humanity’s potential. — Emad
If you don’t want to wait too long in the queue, use the Colab Notebook instead.
Screenshot from Stable Diffusion Colab Notebook
In terms of UI, it’s a little less user-friendly compared to Hugging Faces but is still completely free, and it’s faster since the computing resource is not shared with anyone else.
Now if you want an even more user-friendly UI that also gives you more control over the parameters, use Dream Studio.
Dream Studio
Dream Studio is the paid version of Stable Diffusion that can be accessed here.
When users sign up, they will be given 200 credits for free. But after that, they need to buy credits which costs £1 ($1.18) for every 100 generations. £100 ($117.76) will buy 10,000 generations.
Read my initial tests and thoughts about Dream Studio here.
It’s a permissive license that allows for commercial and non-commercial usage. It focuses on “ethical and legal use of the model as your responsibility and must accompany any distribution of the model.”
Okay, that’s all the exciting news for today. To keep yourself up to date with the latest on Stable Diffusion, join the Discord community here.
If you like to read more stuff like this in Medium, consider supporting me and thousands of other writers by signing up for a membership.