Image generation models have arguably generated the most buzz around the emergence of generative AI in their ability to create seemingly magic results from simple (or more complex) narrative text prompts.

Further, image generation via the most notable applications have probably been the most approachable by “regular” people in that there is no application code or programming involved in creating amazing results.

The most notable image generation applications currently deployed include the following:

DALL-E 2

DALL-E 2 by OpenAI is an AI system that can create realistic images and art from a description in natural language. Find more information about DALL-E 2 here, and you can try it out yourself here (requires a login).

Midjourney

The preferred image generation tool by Tim on our team, information about how to get started with Midjourney can be found here. Midjourney is a little interesting in that you provide your prompts via the Midjourney community on Discord or the Web.

Stable Diffusion

Stable Diffusion XL is a latent text-to-image diffusion model capable of generating photo-realistic images given any text input, cultivates autonomous freedom to produce incredible imagery, empowers billions of people to create stunning art within seconds. More about Stable Diffusion here, including a Playground that is super easy to use.

<aside> ☝ Back to other topics in Explore

</aside>