Is Huihui-GLM-4.7-Flash-abliterated-57B or Qwen3-30B-A3B-Thinking-2507 smaller?

Qwen3-30B-A3B-Thinking-2507 is smaller at Q4_K_M: 17.28 GiB against 31.37 GiB.

Which uses less memory at long context?

Qwen3-30B-A3B-Thinking-2507 — its KV cache at 32K context is 1.4x smaller, because of how its attention layers cache.

Model comparison

Huihui-GLM-4.7-Flash-abliterated-57B vs Qwen3-30B-A3B-Thinking-2507

At Q4_K_M, Qwen3-30B-A3B-Thinking-2507 is the smaller download — 18,556,685,824 bytes against 33,687,658,720. At long context the gap widens: Qwen3-30B-A3B-Thinking-2507's KV cache at 32K is 1.4× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Huihui-GLM-4.7-Flash-abliterated-57B	Qwen3-30B-A3B-Thinking-2507
Parameters	57.3B	30.5B
Architecture	deepseek2	qwen3moe
Layers	63	48
Native context	202,752	262,144
Mixture of experts	yes, 64 experts	yes, 128 experts
Quantizations published	36	51
Smallest quantization	10.79 GiB	7.05 GiB
Q4_K_M	31.37 GiB	17.28 GiB
Licence	—	apache-2.0

KV cache by context

the term that decides long-context viability

Context	Huihui-GLM-4.7-Flash-abliterated-57B	Qwen3-30B-A3B-Thinking-2507	Ratio
4,096	0.52 GiB	0.38 GiB	1.39×
8,192	1.05 GiB	0.75 GiB	1.39×
16,384	2.09 GiB	1.50 GiB	1.39×
32,768	4.18 GiB	3.00 GiB	1.39×
65,536	8.37 GiB	6.00 GiB	1.39×
131,072	16.73 GiB	12.00 GiB	1.39×