The battle of the AI’s!
AI has taken off when it comes to image generation lately and there are a lot of creative things you can do to unlock your imagination and let AI create some pretty cool images for you.
But with the rapid pace of innovation with AI and many, many models coming out what tool will give you the best results?
I put these to the test and ran through a few different scenerios to see what the best is.
The first test I used this prompt:
Portrait of a Shiba Inu dog smiling dressed up as in the American independence war, intricately detailed and realistic. Cinematic, unreal engine, color grading
Here are the results:
Gemini
Grok 1
Grok 2
ChatGPT
MidJourney
Midjourney gives you 4 to choose from so i’ll add them all here
The winner in my opinion is MidJourney – they add a lot more context to the image, the background has a good bokeh effect and overall much more detail compared to the others. Grok 1 was the worst but the Grok 2 seems to have come a long way.
Let’s test out a more straightforward prompt for a realistic photo
create an image of street photo in new york, medium shot, natural lighting, shot on fujifilm, detailed and realistic environment, cinematic
Gemini
Grok2
ChatGPT
MidJourney
Here the prompt is very simple, I’m not giving it a lot of direction to see how the models will each take a different approach. They all have their pros and cons but overall here I think that Gemeni and ChatGpt are the weakest examples.
MidJourney doens’t show a lot of the city but this feels like a NYC street photo.
This isn’t an indepth study or a scientific approach to what LLM is better but hopefully gives you some things to think about.
Which do you prefer and what do you use on a regular basis?
