I've been doing a lot of crazy research with CHATGPT and Gemini, I attempted to do the same with Grok, but Grok was pathetically poor, just like I also stated yesterday. and from the header image in this post you can actually see it.
However before I even start analyse the image creation process, which app do you think go through image creation better?
I'd like to say CHATGPT has a post-generation option that let's you edit the result of your prompt
I mistakenly started using Gemini since it was preinstalled in my Samsung phone and perfectly synchronizes with your Google account, and I was really blown away.
I talked about having a first terrible experience using CHATGPT, and I didn't think CHATGPT could actually match whatever Gemini brought to the table and I was right, but something incredible happened and that's what I'm about to share.
I decided to download the main CHATGPT app on my phone and I saw that it was incredibly better than when you use it on a mobile version. However it's super power is something secret and wait for it. I saw a relatively smoother and better experience with the app.
It offer you the ability to switch between AIs and you can select between other types of prompt option which is great, but this wasn't what blew my mind away.
I still found the performances to be relatively average.. when you switch to deep research, it become very slow and sluggish and annoying, it then lags and crash and actually slows you down, and even while I'm typing, the "deep research" option is still scanning and still not responded and it's been over two minutes for just a simple prompt........ Okay after 9 minutes of deep analysis, it came up with something decent, and I checked it was good, but again, even the "fast" which was the basic Gemini model outperformed this deep research.
The reason is because CHATGPT behaves like a traditional AI (which it is) but Gemini is practical, understanding intent, does better analysis. I compared the results I got and honestly again, the Gemini with only 4 search levels outperformed CHATGPT with 5 or 6.
However this wasn't the same with image generation. Grok's image generation is one of the most appalling I've seen, perhaps I need to try more prompts, but with the few image prompts I put in, it's simply enough to conclude that GROK shouldn't even be here.
Gemini does a very good job, and would have honestly been on par with CHATGPT but that aftereffect which CHATGPT provides at the free level just let's you tweak your output even better and add an extra processing and glow up to your image which is what made CHATGPT take the win when it comes to image creation.
I'll probably need to do a comparison with Claude AI, but Claude currently cannot generate image at its free or basic model.
In conclusion
I tried a few other prompts, asking Gemini to cartoonize an image I fed it, it was good but nothing extraordinary. So this makes CHATGPT the king in the free image category.
The ability to create real life imaginations, although it's not good enough yet for thumbnail creation on places like YouTube, except maybe you're using a paid version (which is what YouTube wants, especially when you're depending on AI to make 100% of your content, from motion to graphs and image generation) but it's the best when it comes to usability and quality.
However Gemini completely murders CHATGPT when it comes to text, research, simplicity and relativity and overall research.