> In-context, your reply reads like you're asserting that anthropic's offerings give you all the same control as a local model, and that having this baseline of control over a computer program is "utopian".

No, I’m talking about the idea up thread that having control over more settings removes all LLM defects with local models.

I have all the control I want over the local models they run. No combination of settings is going to make them outperform Anthropic’s models.