Model comparison

Qwen3.6-27B vs Llama-4-Scout-17B-16E-Instruct-abliterated-v2

These two publish different quantization sets; the table below has the exact sizes.

From the file· summed bytes, KV per layer

Side by side

Qwen3.6-27BLlama-4-Scout-17B-16E-Instruct-abliterated-v2
Parameters27.8B109B
Architectureqwen35llama4
Layers6448
Native context262,14410,485,760
Mixture of expertsnoyes, 16 experts
Quantizations published4015
Smallest quantization8.74 GiB20.66 GiB
Q4_K_M15.41 GiB
Licenceapache-2.0

KV cache by context

the term that decides long-context viability
ContextQwen3.6-27BLlama-4-Scout-17B-16E-Instruct-abliterated-v2Ratio
4,0960.25 GiB
8,1920.50 GiB
16,3841.00 GiB
32,7682.00 GiB
65,5364.00 GiB
131,0728.00 GiB