Is Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP or EXAONE-4.5-33B smaller?

EXAONE-4.5-33B is smaller at Q4_K_M: 18.67 GiB against 34.04 GiB.

Which uses less memory at long context?

Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP — its KV cache at 32K context is 1.4x smaller, because of how its attention layers cache.

Model comparison

Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP vs EXAONE-4.5-33B

At Q4_K_M, EXAONE-4.5-33B is the smaller download — 20,047,839,424 bytes against 36,545,830,912. At long context the gap widens: Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP'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

	Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP	EXAONE-4.5-33B
Parameters	27.8B	34.4B
Architecture	qwen35	exaone4
Layers	64	64
Native context	262,144	262,144
Mixture of experts	no	no
Quantizations published	34	31
Smallest quantization	10.12 GiB	11.57 GiB
Q4_K_M	34.04 GiB	18.67 GiB
Licence	apache-2.0	other

KV cache by context

the term that decides long-context viability

Context	Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP	EXAONE-4.5-33B	Ratio
4,096	0.25 GiB	1.00 GiB	4.00×
8,192	0.50 GiB	1.34 GiB	2.69×
16,384	1.00 GiB	1.84 GiB	1.84×
32,768	2.00 GiB	2.84 GiB	1.42×
65,536	4.00 GiB	4.84 GiB	1.21×
131,072	8.00 GiB	8.84 GiB	1.11×