Model comparison

Qwen3.6-27B vs DeepSeek-MOE-4X8B-R1-Distill-Llama-3.1-Deep-Thinker-Uncensored-24B

At Q4_K_M, DeepSeek-MOE-4X8B-R1-Distill-Llama-3.1-Deep-Thinker-Uncensored-24B is the smaller download — 15,162,181,696 bytes against 16,547,397,888.

From the file· summed bytes, KV per layer

Side by side

Qwen3.6-27BDeepSeek-MOE-4X8B-R1-Distill-Llama-3.1-Deep-Thinker-Uncensored-24B
Parameters27.8B24.9B
Architectureqwen35llama
Layers64
Native context262,144
Mixture of expertsnono
Quantizations published4011
Smallest quantization8.74 GiB8.66 GiB
Q4_K_M15.41 GiB14.12 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextQwen3.6-27BDeepSeek-MOE-4X8B-R1-Distill-Llama-3.1-Deep-Thinker-Uncensored-24BRatio
4,0960.25 GiB
8,1920.50 GiB
16,3841.00 GiB
32,7682.00 GiB
65,5364.00 GiB
131,0728.00 GiB