Model comparison

Llama3.2-30B-A3B-II-Dark-Champion-INSTRUCT-Heretic-Abliterated-Uncensored vs Qwen3.6-27B

At Q4_K_M, Qwen3.6-27B is the smaller download — 16,547,397,888 bytes against 18,366,980,192. At long context the gap widens: Qwen3.6-27B's KV cache at 32K is 2.9× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

Llama3.2-30B-A3B-II-Dark-Champion-INSTRUCT-Heretic-Abliterated-UncensoredQwen3.6-27B
Parameters30.0B27.8B
Architecturellamaqwen35
Layers4764
Native context131,072262,144
Mixture of expertsyes, 8 expertsno
Quantizations published3440
Smallest quantization6.07 GiB8.74 GiB
Q4_K_M17.11 GiB15.41 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextLlama3.2-30B-A3B-II-Dark-Champion-INSTRUCT-Heretic-Abliterated-UncensoredQwen3.6-27BRatio
4,0960.73 GiB0.25 GiB2.94×
8,1921.47 GiB0.50 GiB2.94×
16,3842.94 GiB1.00 GiB2.94×
32,7685.88 GiB2.00 GiB2.94×
65,53611.75 GiB4.00 GiB2.94×
131,07223.50 GiB8.00 GiB2.94×