Model comparison

Qwen3.6-27B vs L3-V2-Stheno-Maid-Blackroot-Grand-HORROR-17B-GLM4.7-Thinking-UNCENSORED

At Q4_K_M, L3-V2-Stheno-Maid-Blackroot-Grand-HORROR-17B-GLM4.7-Thinking-UNCENSORED is the smaller download — 10,142,098,784 bytes against 16,547,397,888.

From the file· summed bytes, KV per layer

Side by side

Qwen3.6-27BL3-V2-Stheno-Maid-Blackroot-Grand-HORROR-17B-GLM4.7-Thinking-UNCENSORED
Parameters27.8B16.5B
Architectureqwen35llama
Layers64
Native context262,144
Mixture of expertsnono
Quantizations published4034
Smallest quantization8.74 GiB3.55 GiB
Q4_K_M15.41 GiB9.45 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextQwen3.6-27BL3-V2-Stheno-Maid-Blackroot-Grand-HORROR-17B-GLM4.7-Thinking-UNCENSOREDRatio
4,0960.25 GiB
8,1920.50 GiB
16,3841.00 GiB
32,7682.00 GiB
65,5364.00 GiB
131,0728.00 GiB