Choose your AI model

Use GPT Sol (medium) as your daily driver.

Which AI model should you use?

Claude Opus 5.5 is the best model right now. It fixes most of Opus 5’s problems, has a great personality, and writes well. But the harness matters just as much as the model. ChatGPT is still far ahead there.

So again, if you could only pick one, go with ChatGPT and GPT Sol. Here’s my take on the current models:

GPT-6 Sol is my daily driver in ChatGPT. Astra is good, but in my opinion, more unreliable than Sol.
Claude Opus 5.5 is the best model for writing, design, and stress-testing important decisions. If you can afford two subscriptions, I would get Claude too since Opus 5.5 really is that good.
Grok is fast and useful for iterating quickly.
Gemini currently lags behind all the options above.

If you use ChatGPT, default to GPT-6 Sol (medium)

On ChatGPT, your model choices are GPT-6 Astra, Sol, and Luna. Each model also has six different effort levels. I know, it’s confusing as hell.

For daily work, GPT-6 Sol (medium) has the best balance between performance and cost. To get the most out of it, you’ll need to connect your work apps.