Model comparison

Llama-3.2-8X3B-GATED-MOE-Reasoning-Dark-Champion-Instruct-uncensored-abliterated-18.4B vs Qwen3-30B-A3B-Thinking-2507

At Q4_K_M, Llama-3.2-8X3B-GATED-MOE-Reasoning-Dark-Champion-Instruct-uncensored-abliterated-18.4B is the smaller download — 11,408,360,768 bytes against 18,556,685,824.

From the file· summed bytes, KV per layer

Side by side

Llama-3.2-8X3B-GATED-MOE-Reasoning-Dark-Champion-Instruct-uncensored-abliterated-18.4BQwen3-30B-A3B-Thinking-2507
Parameters18.4B30.5B
Architecturellamaqwen3moe
Layers48
Native context262,144
Mixture of expertsnoyes, 128 experts
Quantizations published1151
Smallest quantization6.65 GiB7.05 GiB
Q4_K_M10.62 GiB17.28 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextLlama-3.2-8X3B-GATED-MOE-Reasoning-Dark-Champion-Instruct-uncensored-abliterated-18.4BQwen3-30B-A3B-Thinking-2507Ratio
4,0960.38 GiB
8,1920.75 GiB
16,3841.50 GiB
32,7683.00 GiB
65,5366.00 GiB
131,07212.00 GiB