Is Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP or Llama-4-Scout-17B-16E-Instruct smaller?

Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP is smaller at Q4_K_M: 34.04 GiB against 60.87 GiB.

Which uses less memory at long context?

We cannot resolve the KV layout for one of these models, so we do not publish a comparison.

Model comparison

At Q4_K_M, Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP is the smaller download — 36,545,830,912 bytes against 65,359,900,352.

From the file· summed bytes, KV per layer

	Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP	Llama-4-Scout-17B-16E-Instruct
Parameters	27.8B	109B
Architecture	qwen35	llama4
Layers	64	48
Native context	262,144	10,485,760
Mixture of experts	no	yes, 16 experts
Quantizations published	34	48
Smallest quantization	10.12 GiB	24.51 GiB
Q4_K_M	34.04 GiB	60.87 GiB
Licence	apache-2.0	other

the term that decides long-context viability

Context	Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP	Llama-4-Scout-17B-16E-Instruct	Ratio
4,096	0.25 GiB	—	—
8,192	0.50 GiB	—	—
16,384	1.00 GiB	—	—
32,768	2.00 GiB	—	—
65,536	4.00 GiB	—	—
131,072	8.00 GiB	—	—