I never made this journey, but I have ended up with the same AI coding stack.

I've found Cursor to be frustratingly bad recently. I agree with the UX and the over-eager autocompletions. The UI seems to change with every update too. I've had my own problems as well. I can't select any models, being forced to use "auto" permanently. I assume I've used any allowances in my plan, but the app doesn't give any feedback. Chats just fail to start.

Whatever auto is using is spitting out pure gibberish about 30% of the time. What was mostly a huge productivity boost can other times be a massive time sink.

I'm going to give antigravity a go for a while, I've found Gemini to be pretty good in the past.

Thanks for the update - Really wishing all devs would post when they switch IDEs and turn not only their workflow but whole entire paradigm on its head!

I’ll express my personal preference for VS Code with Cline. I don’t know exactly why, but its workflow feels right even though they’re all almost the same interface. I do like the huge choice of models and payment options. For something I use burstily, it makes sense for me to pay as I go.

Is it only me or „Apply” barely works lately? The

To be honest, I'm surprised that people say Claude Code performs THAT much better than Cursor, I found them to be very similar in ability considering that I'm only on Claude Opus 4.5. I prefer cursor because it's easier to go back to previous chats/previous "states".

Also, you CAN choose the same model now in Cursor (as said, I always leave it on Opus 4.5).

Does VSCode have the ability to index docs now? Until then I will continue to use Cursor as that is a fundamental feature necessary if you're using any sort of package that may not necessarily be in the training data. And no, context7 doesn't cut it.

I never got the point of going out with forks instead the real deal, eventually the real deal will get the features that are most relevant, while forks are dependent on where upstream goes.