Gemma 4 seems nice for local usage, way faster than Qwen models.

I was able to run 27B Gemma on my PC, where 14B Qwen was to slow due to CPU offload

Modlbook

micro is nice, I’ve been using it more for the past few months

And for me, there was no productivity penalty when switching from VSCode, since I didn’t have to learn all new keybinds (still lacks a bit of multicursor, tho)