Is Qwen3.5-9B-Neo or Muse-Glimmer-30B smaller?

Qwen3.5-9B-Neo is smaller at Q4_K_M: 5.50 GiB against 16.12 GiB.

Which uses less memory at long context?

Muse-Glimmer-30B — its KV cache at 32K context is 2.0x smaller, because of how its attention layers cache.

Model comparison

Qwen3.5-9B-Neo vs Muse-Glimmer-30B

At Q4_K_M, Qwen3.5-9B-Neo is the smaller download — 5,910,779,968 bytes against 17,306,324,000. At long context the gap widens: Muse-Glimmer-30B'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

	Qwen3.5-9B-Neo	Muse-Glimmer-30B
Parameters	9.7B	29.8B
Architecture	qwen35	muse-glimmer
Layers	32	52
Native context	262,144	131,072
Mixture of experts	no	no
Quantizations published	24	36
Smallest quantization	3.36 GiB	8.31 GiB
Q4_K_M	5.50 GiB	16.12 GiB
Licence	apache-2.0	apache-2.0

KV cache by context

the term that decides long-context viability

Context	Qwen3.5-9B-Neo	Muse-Glimmer-30B	Ratio
4,096	0.13 GiB	0.15 GiB	1.17×
8,192	0.25 GiB	0.20 GiB	1.27×
16,384	0.50 GiB	0.30 GiB	1.68×
32,768	1.00 GiB	0.50 GiB	1.99×
65,536	2.00 GiB	0.91 GiB	2.20×
131,072	4.00 GiB	1.72 GiB	2.33×