Model comparison

Qwen3.6-27B vs Llama3.2-24B-A3B-II-Dark-Champion-INSTRUCT-Heretic-Abliterated-Uncensored

At Q4_K_M, Llama3.2-24B-A3B-II-Dark-Champion-INSTRUCT-Heretic-Abliterated-Uncensored is the smaller download — 11,091,316,864 bytes against 16,547,397,888. At long context the gap widens: Qwen3.6-27B's KV cache at 32K is 1.7× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

Qwen3.6-27BLlama3.2-24B-A3B-II-Dark-Champion-INSTRUCT-Heretic-Abliterated-Uncensored
Parameters27.8B18.0B
Architectureqwen35llama
Layers6428
Native context262,144131,072
Mixture of expertsnoyes, 8 experts
Quantizations published4035
Smallest quantization8.74 GiB3.72 GiB
Q4_K_M15.41 GiB10.33 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextQwen3.6-27BLlama3.2-24B-A3B-II-Dark-Champion-INSTRUCT-Heretic-Abliterated-UncensoredRatio
4,0960.25 GiB0.44 GiB1.75×
8,1920.50 GiB0.88 GiB1.75×
16,3841.00 GiB1.75 GiB1.75×
32,7682.00 GiB3.50 GiB1.75×
65,5364.00 GiB7.00 GiB1.75×
131,0728.00 GiB14.00 GiB1.75×
Qwen3.6-27B vs Llama3.2-24B-A3B-II-Dark-Champion-INSTRUCT-Heretic-Abliterated-Uncensored — size, memory and hardware fit — ossmodeldb