From my personal testing, running various agentic tasks with a bunch of tool calls on an M4 Max 128GB, I've found that running quantized versions of larger mode

From my personal testing, running various agentic tasks with a bunch of tool calls on an M4 Max 128GB, I've found that running quantized versions of larger models to produce the best results which this site completely ignores. Currently, Nemotron 3 Super using Unsloth's UD Q4_K_XL quant is running nearly everything I do locally (replacing Qwen3.5 122b)