Is Qwen3-Coder-30B-A3B-Instruct or internlm2-math-plus-20b smaller?

internlm2-math-plus-20b is smaller at Q4_K_M: 11.16 GiB against 17.28 GiB.

Which uses less memory at long context?

Qwen3-Coder-30B-A3B-Instruct — its KV cache at 32K context is 2.0x smaller, because of how its attention layers cache.

Model comparison

Qwen3-Coder-30B-A3B-Instruct vs internlm2-math-plus-20b

At Q4_K_M, internlm2-math-plus-20b is the smaller download — 11,984,469,440 bytes against 18,556,689,568. At long context the gap widens: Qwen3-Coder-30B-A3B-Instruct's KV cache at 32K is 2.0× 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-Instruct	internlm2-math-plus-20b
Parameters	30.5B	19.9B
Architecture	qwen3moe	internlm2
Layers	48	48
Native context	262,144	8,192
Mixture of experts	yes, 128 experts	no
Quantizations published	46	68
Smallest quantization	7.46 GiB	4.23 GiB
Q4_K_M	17.28 GiB	11.16 GiB
Licence	apache-2.0	other

KV cache by context

the term that decides long-context viability

Context	Qwen3-Coder-30B-A3B-Instruct	internlm2-math-plus-20b	Ratio
4,096	0.38 GiB	0.75 GiB	2.00×
8,192	0.75 GiB	1.50 GiB	2.00×
16,384	1.50 GiB	3.00 GiB	2.00×
32,768	3.00 GiB	6.00 GiB	2.00×
65,536	6.00 GiB	12.00 GiB	2.00×
131,072	12.00 GiB	24.00 GiB	2.00×