How and Where to Begin Creating AI Images with Text-to-Image Generators? Your First AI Image! Stable Diffusion XL (SDXL), DALL-E 3, Midjourney
If you’ve never tried creating AI-generated images before, I warn you that starting this venture will suddenly change your life for the better!:-)
It’s an incredibly creative and engaging activity suitable for individuals of all ages. As a child, I ardently desired to draw and paint, but unfortunately lacked the talent. AI has allowed me to fulfill those dreams! Today, I can draw anything in any style and technique.
For half of my life, I’ve been professionally involved in photography, which also provides a creative outlet. However, AI opens up far greater possibilities, enabling me to create photos that would be impossible in the real world.
In just a few months, I’ve learned more about art than in my entire life. The same will happen to you! You’ll discover entirely new ways of expressing yourself. You’ll become a digital artist, and my tutorials will assist you on this journey:
I could go on about the advantages of creating AI images, but I know you’re eager to create your first artwork, so let’s get to it!
Currently, three image generation models dominate the market: Midjourney, DALL-E 3, and Stable Diffusion XL. Each has its pros and cons, so it’s worth trying them all to find your favorite.
Midjourney
I don’t recommend the Midjourney model for beginners because it’s not available for free and is too complex to handle. It can be an excellent choice once you gain experience.
DALL-E 3
The DALL-E 3 model is available in Chat-GPT (in the paid option) and on the Bing search engine (for free). The easiest way is to use the Bing Image Generator. Simply create an account on Bing, and the image generator is accessible here: www.bing.com/images/create
It supports over 100 languages, so just enter a prompt in the text box, describing what you want to create. For example, a prompt could be:
portrait of a cute kitten in the garden, intricate details
One window, simple rules, so no specialized knowledge is needed. After a moment, you’ll receive a set of four images with a resolution of 1024x1024 pixels.
Simplicity is a great advantage, but it can also be a drawback because you’ll quickly want to create AI images with more control and explore other possibilities offered by text-to-image services.
Stable Diffusion XL
This is my favorite model! In my tutorials, I primarily use the SDXL model because it produces high-quality images, has immense capabilities, and evolves rapidly. Stable Diffusion is free and open-source, allowing anyone to run it locally if they have a powerful enough computer.
This makes Stable Diffusion incredibly popular and available on many platforms, including free options!
So, after a brief exploration of DALL-E 3, let’s move on to services offering the SDXL model. Best in the following order:
1) Ionium — fumesai.web.app/sdxl.html
Without logging in and free of charge, you can immediately enter prompts and create images. It has simple additional options that you don’t need to change. It’s great for checking how prompts work. During peak hours, it may work slowly.
I don’t know how long it will remain available without logging in and for free, so it might change quickly.
2) Playground AI — playground.com
This is already a professional service with the SDXL model, but it’s still easy to use. And most importantly, you can generate up to 500 images for free every day and use them for commercial purposes, meaning you can earn from them. This is probably the best free offer on the market.
领英推荐
Currently, the only way to create an account and log in to the service is by using the Google login option. This means you need to have a Google account to log in to Playground.
Illustrations can be created in two ways: using the “Board” option (an easy way to generate art) and “Canvas” (a powerful image editor). Start by creating images on the board. You can go there through the menu or by clicking the link: playgroundai.com/create
The three parameters that interest us the most are:
a) “Prompt” on the left side is the most important area. In this field, enter prompts, which are commands describing how our image should look. Now, enter a prompt to create a kitten:
portrait of a cute kitten in the garden, intricate details
b) “Model” on the right side at the top specifies the AI model for generating the image. Set it to Stable Diffusion XL — it produces excellent results.
c) “Image Dimension” is the field defining the size of the image. Set the resolution to the highest, which is 1024x1024.
You can also check “Private Sessions” in the bottom right corner (scroll the right panel as this setting is at the bottom). The images will then be visible only to you and won’t be publicly accessible.
Now, click the “Generate” button located on the left at the bottom. After a few seconds, the kitten will be generated!
Congratulations, you now know how to create AI images!
You can click the “Generate” button again to generate a new image for the given prompt, or you can change the prompt to a different one.
You can save the image to your disk, edit it, make it public, or delete it. Every generated image goes to your profile, where you can review its parameters and prompts. On the profile, you can specify whether the image will be publicly visible or remain private.
For most options, you’ll find short descriptions of their functions and even instructional videos. Take some time to explore and experiment with various settings.
3) NightCafe — creator.nightcafe.studio
Playground AI is excellent for working with AI images, but if you enjoy communities and interacting with other users, commenting on works, and participating in contests, NightCafe is unparalleled.
However, it’s considerably more expensive and slightly more complex to use. The principles of creating images are similar everywhere, so once you master using one service, you’ll manage others, and then you can choose your favorite.
I have paid accounts on Playground AI and NightCafe, so I actively use both.
What’s next?
You’ll find more essential and useful information in my tutorials:
You’ll learn how to create extraordinary images, build effective prompts, use various painting techniques and styles, or how to create even better illustrations inspired by the styles of the best artists. I’ll also teach you how to create images for commercial purposes and earn from them in many ways.
As you can see, creating images is not complicated. In one afternoon, you can start the adventure of a lifetime! Congratulations on becoming an AI artist, and I wish you success on this new artistic path!:-)
And be sure to share the passion for digital AI art with others by sending them the link to this tutorial. Let them start creating and join us!