AI Art: Comparing Generators and Their Differences

Artificial intelligence art creation is going through one of its best moments. With just a description, algorithms can shape incredible portraits, landscapes, and situations, applying special styles and delivering photorealistic results in seconds. Fabian Stelzer on Twitter decided to organize a comparison of the main generators, a true head-to-head between DALL-E 2 and the two most popular alternatives, Midjourney and Stable Diffusion.

One sentence, some extra details, and a click. That's all it takes these days to create impressive portraits and landscapes using artificial intelligence. Imagination seems to be the limit, but in reality it depends on the training that the algorithm in question has received. At the same time, that introduces variables and secondary adjustments that produce completely different results, even when the tools' goal is the same.

DALL-E 2, Midjourney, and Stable Diffusion are three art generation models that have taken the Web by storm. They are so flexible and powerful that they have been declared as "dream tools" and "industry destroyers" alike. So... why not pit them against each other? Fabian Stelzer did exactly that on Twitter, applying the same rules in all cases: identical description, 1:1 aspect ratio, and no living artists. Below you can enjoy a quick selection, but I definitely recommend exploring the rest of the thread on Twitter. Who do you think will win?

AI Art Generator Showdown

Source: Fabian Stelzer on Twitter