Is Qwen3.6-27B or Wizard-Vicuna-30B-Uncensored smaller?

Qwen3.6-27B is smaller at Q4_K_M: 15.41 GiB against 18.27 GiB.

Which uses less memory at long context?

Qwen3.6-27B — its KV cache at 32K context is 24.4x smaller, because of how its attention layers cache.

Model comparison

Qwen3.6-27B vs Wizard-Vicuna-30B-Uncensored

At Q4_K_M, Qwen3.6-27B is the smaller download — 16,547,397,888 bytes against 19,621,139,904. At long context the gap widens: Qwen3.6-27B's KV cache at 32K is 24.4× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Qwen3.6-27B	Wizard-Vicuna-30B-Uncensored
Parameters	27.8B	32.5B
Architecture	qwen35	llama
Layers	64	60
Native context	262,144	2,048
Mixture of experts	no	no
Quantizations published	40	33
Smallest quantization	8.74 GiB	6.63 GiB
Q4_K_M	15.41 GiB	18.27 GiB
Licence	apache-2.0	other

KV cache by context

the term that decides long-context viability

Context	Qwen3.6-27B	Wizard-Vicuna-30B-Uncensored	Ratio
4,096	0.25 GiB	6.09 GiB	24.36×
8,192	0.50 GiB	12.19 GiB	24.37×
16,384	1.00 GiB	24.38 GiB	24.37×
32,768	2.00 GiB	48.75 GiB	24.37×
65,536	4.00 GiB	97.50 GiB	24.37×
131,072	8.00 GiB	195.00 GiB	24.37×