Model comparison

Qwen3-Coder-30B-A3B-Instruct vs Omega-Evolution-26B-A4B-v3.0

At Q4_K_M, Qwen3-Coder-30B-A3B-Instruct is the smaller download — 18,556,689,568 bytes against 51,258,890,720.

From the file· summed bytes, KV per layer

Side by side

Qwen3-Coder-30B-A3B-InstructOmega-Evolution-26B-A4B-v3.0
Parameters30.5B26.5B
Architectureqwen3moegemma4
Layers48
Native context262,144
Mixture of expertsyes, 128 expertsno
Quantizations published465
Smallest quantization7.46 GiB13.10 GiB
Q4_K_M17.28 GiB47.74 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextQwen3-Coder-30B-A3B-InstructOmega-Evolution-26B-A4B-v3.0Ratio
4,0960.38 GiB
8,1920.75 GiB
16,3841.50 GiB
32,7683.00 GiB
65,5366.00 GiB
131,07212.00 GiB