How to install and use an artificial intelligence on your PC to generate images

Stable Diffusion is an AI that can create images from a description and you can use it locally on your computer

Image-generating artificial intelligences are in for a sweet spot . The success of DALL-E and its latest version, DALL-E 2 , have been joined by others such as Midjourney or Stable Diffusion , the one we are dealing with, and which are revolutionizing the world of image creation.

Not all image-generating AIs have the same accessibility. Some require requesting their use and entering a waiting list while others can be used via the web without further delay. However, Stable Diffusion, in addition to this last option, can also be downloaded and installed on a PC to use it locally on any computer and without the need to connect to the Internet .

Stable Diffusion requires Windows 10 or 11 and an Nvidia graphics card with at least 6GB of VRAM, but if the computer has one from AMD or Intel, the program can work by doing the necessary calculations with the CPU , although it will be a slower process. By slow we mean quadrupling the time required, but this will vary depending on the CPU and GPU used in each case.

In order to use Stable Diffusion on your PC, you must first access the GitHub repository where the tool is located and proceed to download it. It is as simple as choosing the option for Windows, there is also one for Linux. Once the compressed file of just over 200 MB has been downloaded with Stable Diffusion, you must follow these steps:

  • Unzip the downloaded stable-diffusion-ui file .
  • Copy the unzipped folder and paste it in the root directory of your computer, usually “C:” The path to the folder should be C:\stable-diffusion-ui . This way you will avoid problems with character limits in file paths.
  • Run the Start Stable Diffusion UI.cmd file found in the folder you copied by double-clicking on it. This will open the Windows Command Console , where you will be prompted to press any key to start the process.
  • The program will download the remaining software as it completes the installation of Stable Diffusion. The process is long and can perfectly reach a quarter of an hour, depending on the equipment and connection speed. It is possible that the Windows firewall asks you for permission at some point to allow connections from one of its components, which you must confirm.
  • Once Stable Diffusion completes the installation, your browser will open a web address showing the interface for using the tool: http://localhost:9000/. Every time you want to start the AI ​​you will have to run the file Start Stable Diffusion UI.cmd , but the start will be immediate without having to repeat the download process. You can create a shortcut on the Desktop to have it more at hand.

The interface of the tool is extremely simple. The space for the prompt is where you should write the description of the image you want to create. You can also upload an image from which to create the new one.

In the Advanced Settings section you can set options such as image resolution, up to 2K with the possibility of scaling it to 4K, or the number of images to generate. You can also introduce modifiers such as the visual or drawing style, rendering styles, the type of camera and point of view, color palettes or even apply the style of classic artists from painting, comics or cinema, among many others. options, to the prompt entered. The possibilities with Stable Diffusion are up to the user’s imagination.

Previous articleBrave will allow you to block cookie notices from next month
Next articleEleven keys to evolve into a Protected Organization