With 64GB of RAM you should look into Qwen3.5-27B or Qwen3.5-35B-A3B. I suggest Q5 quantization at most from my experience. Q4 works on short responses but gets weird in longer conversations.
There are dynamic quants such as Unsloth which quantize only certain layers to Q4. Some layers are more sensitive to quantization than others. Smaller models are more sensitive to quantization than the larger ones. There are also different quantization algorithms, with different levels of degradation. So I think it's somewhat wrong to put "Q4" under one umbrella. It all depends.
Nobody actually quantizes every layer to Q4 in a Q4 quant.
This approach also makes less sense for discrete GPUs where VRAM is quite fast but scarce, and the GPU's PCIe link is a key bottleneck. I suppose it starts to make sense again once you're running the expert layers with CPU+RAM.
Also I wouldn’t trust 3-bit quantization for anything real. I run a 5-bit qwen3.5-35b-A3B MoE model on my studio for coding tasks and even the 4-bit quant was more flaky (hallucinations, and sometimes it would think about running tools calls and just not run them, lol).
If you decided to give it a go make sure to use the MLX over the GGUF version! You’ll get a bit more speed out of it.