I would normally never upgrade so soon after getting an M1 but running local LLMs is extremely cool and useful to the point where I'd want the extra RAM and CPU to run larger models more quickly.
I'm bumping from a still-excellent M1 MAX / 64GB to M4 MAX / 128GB, mostly for local GenAI. It gives me some other uplift and also enables me to sell this system while it's still attractive. I'm able to exhaust local 7B models fairly easily on it.