These days, I bounce between ChatGPT, Claude, and Perplexity. The lead tool has changed a few times (along with where I spend my subscription money) - and that's a good thing. Let me explain.
Like most folks, my AI adventures started with ChatGPT. But as time went on, I found myself using Claude more and more, mostly for its better reasoning, content generation and other features. So when I decided to start vibe coding, I went with Claude Code. I also used Claude to build my home AI box, and that process was fairly smooth.
As time passed and I went deeper with Claude, it started to let me down. Claude could do impressive things, but understanding what it had done and getting to a useful result took more work and time than I wanted.
Then I read about someone creating a repo to translate “Claudish” into plain English. That felt like a sign it might be time to try something else.
I’d kept subscriptions to all three, but Claude was getting $200 a month so I could do more coding and run scheduled jobs. When I gave ChatGPT and Codex another try, the difference was immediate. When Codex reported back, I understood what it was saying without needing a translator. And I spent less time messing around, burning tokens and trying to follow the explanation, allowing for more time getting tangible results.
So I moved the money around. I now pay $20 a month for Claude, $20 for Perplexity, and $100 for ChatGPT.
Perplexity has stayed in the mix throughout. I use it for web research and fact-checking the other two hoodlums, and I’ll likely keep that mix going. For everyday building, reasoning, and strategy, ChatGPT and Codex are getting most of my work right now.
But I’m keeping Claude around. It was the better fit for me before, and I’m open to switching again.
That’s how I’d approach choosing between these tools: try more than one on the work you need done. If you’re spending too much time trying to get somewhere with one, stop and give another a shot. That switch has made a huge difference for me, but your mileage may vary.