Is ZAYA1-8B or Qwen3-30B-A3B-Thinking-2507 smaller?

ZAYA1-8B is smaller at Q4_K_M: 5.19 GiB against 17.28 GiB.

Which uses less memory at long context?

ZAYA1-8B — its KV cache at 32K context is 78643.2x smaller, because of how its attention layers cache.

Model comparison

ZAYA1-8B vs Qwen3-30B-A3B-Thinking-2507

At Q4_K_M, ZAYA1-8B is the smaller download — 5,567,581,549 bytes against 18,556,685,824. At long context the gap widens: ZAYA1-8B's KV cache at 32K is 78643.2× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	ZAYA1-8B	Qwen3-30B-A3B-Thinking-2507
Parameters	8.8B	30.5B
Architecture	zaya	qwen3moe
Layers	40	48
Native context	131,072	262,144
Mixture of experts	yes, 16 experts	yes, 128 experts
Quantizations published	6	51
Smallest quantization	4.20 GiB	7.05 GiB
Q4_K_M	5.19 GiB	17.28 GiB
Licence	—	apache-2.0

KV cache by context

the term that decides long-context viability

Context	Qwen3-30B-A3B-Thinking-2507	Ratio
4,096	0.38 GiB	9830.40×
8,192	0.75 GiB	19660.80×
16,384	1.50 GiB	39321.60×
32,768	3.00 GiB	78643.20×
65,536	6.00 GiB	157286.40×
131,072	12.00 GiB	314572.80×