Beautiful Minds: ChatGPT 4.5 vs. Claude 3.7 in an AI Prompt Showdown: Which Model is Best for AI Building Tools?

ChatGPT 4.5 dropped today! I gave it a test drive by asking it to build me a prompt for an AI building tool, and then I tested it against 4 other models include Claude 3.7 to see which prompts best!

We live in a world with many different beautiful artificial minds. If we want to make things, part of the task is understanding which model (mind) to use for which task. This post is an attempt to figure that out for AI assisted building tools, which is a huge category of LLM usage these days.

If you want to know which model I recommend, do a quick doom scroll to the end, but I think the journey is worth it here. My goal is to get you thinking outside the new is better box, or even outside the reasoning is better box: the work that a model does is very dependent on the kind of tuning it has had and I am finding in practice there isn’t a substitute for this kind of detailed work to discover model competencies.

I’ll also add that none of these models are bad. So don’t feel terrible if you have to pick one of the models here and not another. I include tips to help you prompt each model more effectively for AI building tools!