After exploring ChatGPT, Gemini and Claude in the LLM space, I have decided to stick to one, Claude. As of me writing this all the 3 LLM’s have released newer and better versions. To me Claude does things natively which you have to prompt ChatGPT/Gemini. For me, Claude seems to understand my intent better and provide visual results intuitively.
Here is an example. Recently I had a personal request to come up with a draft home plan for a house. Here is my prompt and the response from Claude. If you notice, Claude created a image as well as explained all the key features. You can easily see it was not correct the first time, but I was able to get it tweaked by asking Claude to fix it errors.
For the same prompt with Gemini, here is what I received as response
Raasukutty's 