Model comparison

Qwen3-Coder-30B-A3B-Instruct vs functionary-medium-v3.2

At Q4_K_M, Qwen3-Coder-30B-A3B-Instruct is the smaller download — 18,556,689,568 bytes against 42,520,394,592.

From the file· summed bytes, KV per layer

Side by side

Qwen3-Coder-30B-A3B-Instructfunctionary-medium-v3.2
Parameters30.5B70.6B
Architectureqwen3moellama
Layers4880
Native context262,144131,072
Mixture of expertsyes, 128 expertsno
Quantizations published4618
Smallest quantization7.46 GiB15.60 GiB
Q4_K_M17.28 GiB39.60 GiB
Licenceapache-2.0mit

KV cache by context

the term that decides long-context viability
ContextQwen3-Coder-30B-A3B-Instructfunctionary-medium-v3.2Ratio
4,0960.38 GiB
8,1920.75 GiB
16,3841.50 GiB
32,7683.00 GiB
65,5366.00 GiB
131,07212.00 GiB