Dall-E 3 is the most recent and lately launched AI picture generator from OpenAI, and it has been making waves within the AI house due to its spectacular capabilities. It is a major enchancment over the earlier model of Dall-E, with higher coherence and picture high quality.
But how does Dall-E 3 evaluate to its closest rivals, MidJourney 5.2 and Stable Diffusion XL?
In this text, I’ll evaluate the three fashions utilizing six totally different prompts to see which one produces the perfect outcomes.
1. Coherence
Prompt: An astronaut driving a metal horse on the moon. The astronaut is carrying a medieval armor with a celebration hat and a inexperienced sword.
Coherence is among the highlights of Dall-E 3’s new options. Dall-E 3 nailed it, getting each bizarre element. The others? They sort of forgot the occasion temper (missed the hat) and the sword.
2. Surreal Landscape
Prompt: A dreamy panorama the place clouds are manufactured from cotton sweet and rivers move with liquid gold
Here, I needed to see if these AI fashions might dream somewhat and blend issues up. It was all about turning the loopy into “whoa, that’s fairly.”
The approach they give the impression of being, I’d say Dall-E 3 nonetheless seems the perfect among the many three. By requesting a mixture of recognizable components in unfamiliar methods, you possibly can gauge how effectively every AI understands and constructs scenes from scratch.
3. Historical Fiction
Prompt: Ancient Egyptian pharaoh utilizing a contemporary smartphone whereas driving a dinosaur.
Okay, MidJourney’s end result seems the good on this instance, however it didn’t…
