On the 19th of last month, Stability AI's open source large language model StableLM, the Alpha version of the model has 3 billion and 7 billion parameters, and supports commercial use.

In the past only one month, Stability AI announced again that it will open StableStudio , an AI image generation platform ! It's really a big move once a month...

As an open source implementation of DreamStudio, StableStudio can complete AI image generation, editing, restoration and other work online.

In the future, new features such as AI chatbots, plug-in systems, desktop applications, WebGPU local inference, and ControlNet tools will be added!

GitHub: https://github.com/Stability-AI/StableStudio

As I mentioned in the article before, Emad Mostaque, the founder of Stability AI, has been adhering to the concept of open source and openness since he left the OpenAI team in 2020, and has continued to launch a variety of AI open source models.

Among them, the most influential one is the well-known text generation image model Stable Diffusion.

The continuous open source in recent months has also made many netizens call out: You are the real OpenAI!

The installation and use of this project are also relatively simple. The specific installation process is described below.

install & use

First, clone the project locally and enter the folder:

git clone https://github.com/Stability-AI/StableStudio.git

cd StableStudio

Then install the yarn dependencies and run the local server:

yarn

yarn dev

After the project runs successfully, visit the address on the webpage (the default port is 3000): localhost:3000

Then, you can see the interface as shown below:

At this point, it proves that the project has been successfully deployed.

If you want to use the AI drawing function, you need to configure the DreamStudio API Key.

DreamStudio is an AI image generation platform under Stability AI, and each newly registered account will give 25 points.

PS: The price of 1000 credits is 10 knives, which can generate about 5000 images,

DreamStudio account registration is also relatively simple, you can directly use Google account authorization to log in:

Login address: https://dreamstudio.ai/generate

After successfully registering and logging in, go to the settings page to copy the API Key:

https://dreamstudio.ai/account

After the copy is successful, go to the StableStudio settings page and fill in the Key in the input box at the bottom of the page:

http://localhost:3000/settings

Then, you can fill in the prompt words in the Prompt input box and use StableStudio happily.

The image supports various styles such as animation, movie, punk, sci-fi, pixel art, etc. Click the Style button to view it.

StableStudio uses the latest image generation model Stabile Diffusion XL (SDXL) by default, a latest image generation model for enterprise customers, good at generating more realistic photo-level images.

SDXL is capable of producing more detailed images and compositions than its predecessor, Stable Diffusion 2.1, and includes the following features:

• Better photorealism;
• Enhanced image synthesis and face generation;
• Rich visuals and stunning aesthetics;
• Use shorter prompts to create descriptive images;
• Greater ability to generate clear text;

In addition to generating images based on text, SDXL also supports functions such as image generation, image restoration, and image style extension.

write at the end

The advancement of AI painting technology has allowed us to see many amazing results, such as people who do not need any painting background to create high-quality works of art, or can automatically create works of art based on photos provided by users, and so on.

It can be seen that the application of AI painting technology will become more and more extensive in the future, not only in the field of artistic creation, but also in many other fields, such as design, interior decoration and so on.

With the opening of StableStudio, the cost of building an AI painting platform in the future will drop sharply, which will provide opportunities for more technology companies, individual creators, and art institutions to promote the application and development of AI painting technology.

Everyone can easily use the AI painting platform to create, and can present their own ideas and ideas through AI paintings. This will be an era full of creativity and imagination.

In addition, with the advancement of AI painting technology, more AI painting tools and software will emerge, which will greatly improve the efficiency and quality of AI painting, and provide more people with more convenient, efficient, and low-cost Cost of AI painting services.