Model comparison

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

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

Side by side

Llama-4-Scout-17B-16E-InstructQwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP
Parameters109B27.8B
Architecturellama4qwen35
Layers4864
Native context10,485,760262,144
Mixture of expertsyes, 16 expertsno
Quantizations published4834
Smallest quantization24.51 GiB10.12 GiB
Q4_K_M60.87 GiB34.04 GiB
Licenceotherapache-2.0

KV cache by context

the term that decides long-context viability
ContextLlama-4-Scout-17B-16E-InstructQwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTPRatio
4,0960.25 GiB
8,1920.50 GiB
16,3841.00 GiB
32,7682.00 GiB
65,5364.00 GiB
131,0728.00 GiB