Personally, I would always max out the RAM you can fit into your budget. You might get lower bandwidth (= slower generation) than you do on a Mac if you choose a Strix Halo or DGX Spark, but there are always new tweaks being discovered to speed things up. That being said, with 32GB you should be able to fit an ok quant of 35B-A3B or 27B with some context, with 64GB you should be golden.
i have issues on a m5/64g with 35b-a3b (mlx) it eventually hits a memory cap around 52gb... but i'm pretty happy with `Qwen3.6-27B-Claude-Opus-Reasoning-Distilled-mlx-8Bit`
I'm sure there will be a fix for it, but it illustrates an important broader point I should probably have made above: if you opt for local AI today, expect to run into some issues. Expect to learn a bit about the tools you're using, the not-so-fun way. I'm not recommending it to non-technical friends (yet).