Comment by eru

> First, I get exhausted reading LLM prose. I really don't want to read anything generated by something like Opus 5 at this point.

Agreed.

I find Opus 5, and even Fable, to be overly wordy in eg PR descriptions and code comments.

However, I suspect that's more to do with what they are trained to do by default than LLMs in general. I have a little setup where I tell Claude to work together with Codex to tighten up prose and comments, and for me that produces much more palatable text that needs less human editing afterwards.