Have you ever tried so called open weight models such as Kimi, GLM or Mistral? I really like MiniMax M3 (though I use it more for orchestrating than actual coding)
I probably should but haven’t played around too much. Claude usually does a good enough job. I spend the majority of my time thinking about what needs to be built at a high level, and then it executes for the most part. I’m a bit of a laggard, if enough peers start telling me they are using something else that’s what usually forces me to switch. Just thinking back on my transition so far from manual->typing into ChatGPT->cursor->claude
1