Model comparison

Qwen3-Coder-30B-A3B-Instruct vs Kimi-VL-A3B-Instruct

At Q4_K_M, Kimi-VL-A3B-Instruct is the smaller download — 10,540,747,872 bytes against 18,556,689,568. At long context the gap widens: Kimi-VL-A3B-Instruct's KV cache at 32K is 3.2× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

Qwen3-Coder-30B-A3B-InstructKimi-VL-A3B-Instruct
Parameters30.5B16.4B
Architectureqwen3moedeepseek2
Layers4827
Native context262,144131,072
Mixture of expertsyes, 128 expertsyes, 64 experts
Quantizations published4639
Smallest quantization7.46 GiB4.78 GiB
Q4_K_M17.28 GiB9.82 GiB
Licenceapache-2.0mit

KV cache by context

the term that decides long-context viability
ContextQwen3-Coder-30B-A3B-InstructKimi-VL-A3B-InstructRatio
4,0960.38 GiB0.12 GiB3.16×
8,1920.75 GiB0.24 GiB3.16×
16,3841.50 GiB0.47 GiB3.16×
32,7683.00 GiB0.95 GiB3.16×
65,5366.00 GiB1.90 GiB3.16×
131,07212.00 GiB3.80 GiB3.16×
Qwen3-Coder-30B-A3B-Instruct vs Kimi-VL-A3B-Instruct — size, memory and hardware fit — ossmodeldb