Model comparison

Mistral-Nemo-12B-Instruct vs Qwen3.6-27B

At Q4_K_M, Mistral-Nemo-12B-Instruct is the smaller download — 7,477,208,320 bytes against 16,547,397,888.

From the file· summed bytes, KV per layer

Side by side

Mistral-Nemo-12B-InstructQwen3.6-27B
Parameters12.2B27.8B
Architecturellamaqwen35
Layers64
Native context262,144
Mixture of expertsnono
Quantizations published1140
Smallest quantization4.46 GiB8.74 GiB
Q4_K_M6.96 GiB15.41 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextMistral-Nemo-12B-InstructQwen3.6-27BRatio
4,0960.25 GiB
8,1920.50 GiB
16,3841.00 GiB
32,7682.00 GiB
65,5364.00 GiB
131,0728.00 GiB