Is Qwen3-Coder-30B-A3B-Instruct or gemma-3-27b-it-abliterated-refined-vision smaller?

gemma-3-27b-it-abliterated-refined-vision is smaller at Q4_K_M: 15.41 GiB against 17.28 GiB.

Which uses less memory at long context?

Qwen3-Coder-30B-A3B-Instruct — its KV cache at 32K context is 1.0x smaller, because of how its attention layers cache.

Model comparison

At Q4_K_M, gemma-3-27b-it-abliterated-refined-vision is the smaller download — 16,546,689,504 bytes against 18,556,689,568.

From the file· summed bytes, KV per layer

	Qwen3-Coder-30B-A3B-Instruct	gemma-3-27b-it-abliterated-refined-vision
Parameters	30.5B	27.4B
Architecture	qwen3moe	gemma3
Layers	48	62
Native context	262,144	131,072
Mixture of experts	yes, 128 experts	no
Quantizations published	46	34
Smallest quantization	7.46 GiB	5.83 GiB
Q4_K_M	17.28 GiB	15.41 GiB
Licence	apache-2.0	gemma

the term that decides long-context viability

Context	Qwen3-Coder-30B-A3B-Instruct	gemma-3-27b-it-abliterated-refined-vision	Ratio
4,096	0.38 GiB	0.92 GiB	2.46×
8,192	0.75 GiB	1.23 GiB	1.65×
16,384	1.50 GiB	1.86 GiB	1.24×
32,768	3.00 GiB	3.11 GiB	1.04×
65,536	6.00 GiB	5.61 GiB	1.07×
131,072	12.00 GiB	10.61 GiB	1.13×