When do you find Claude better than ChatGPT, and vice versa? Coding vs personal discussion vs general use

I'm trying to figure out where Claude and ChatGPT each work best for me—particularly for coding, personal reflection/discussion, and more general tasks like finding a restaurant or helping make a decision. For coding , both seem pretty capable. I've been using Claude Code through VS Code and have been impressed with it. I installed it first and honestly didn't realize Codex could be used similarly, so I haven't really given Codex a fair comparison yet. For personal reflection/discussion , I tend to like ChatGPT's verbosity. It doesn't always immediately get the point I'm trying to make, but it tends to spell out its reasoning and explore different angles, which can actually be useful. Claude can sometimes be more concise and poignant, which I appreciate, but I've also found that it can become surprisingly judgmental or invoke safeguards in ways that make it harder to simply think through something. For general assistant-type use —travel, restaurants, purchases, random questions, etc.—I'm less sure. Claude seems more willing or able to bring its memory of previous conversations into the discussion, which can make it feel more context-aware. On the other hand, ChatGPT's more detailed answers can be genuinely useful when I'm trying to work through a decision rather than just get a quick answer. For people who use both regularly, where do you think each one currently has the advantage? I'm especially interested in whether you've settled into using one for coding and the other for conversation/general assistant tasks.