Is Ministral-3-8B-Reasoning-2512-heretic or Qwen3.5-9B smaller?

Ministral-3-8B-Reasoning-2512-heretic is smaller at Q4_K_M: 4.84 GiB against 5.24 GiB.

Which uses less memory at long context?

Qwen3.5-9B — its KV cache at 32K context is 4.2x smaller, because of how its attention layers cache.

Model comparison

Ministral-3-8B-Reasoning-2512-heretic vs Qwen3.5-9B

At Q4_K_M, Ministral-3-8B-Reasoning-2512-heretic is the smaller download — 5,198,385,440 bytes against 5,627,044,256. At long context the gap widens: Qwen3.5-9B's KV cache at 32K is 4.2× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Ministral-3-8B-Reasoning-2512-heretic	Qwen3.5-9B
Parameters	8.9B	9.7B
Architecture	mistral3	qwen35
Layers	34	32
Native context	262,144	262,144
Mixture of experts	no	no
Quantizations published	3	34
Smallest quantization	4.84 GiB	2.97 GiB
Q4_K_M	4.84 GiB	5.24 GiB
Licence	apache-2.0	apache-2.0

KV cache by context

the term that decides long-context viability

Context	Ministral-3-8B-Reasoning-2512-heretic	Qwen3.5-9B	Ratio
4,096	0.53 GiB	0.13 GiB	4.25×
8,192	1.06 GiB	0.25 GiB	4.25×
16,384	2.13 GiB	0.50 GiB	4.25×
32,768	4.25 GiB	1.00 GiB	4.25×
65,536	8.50 GiB	2.00 GiB	4.25×
131,072	17.00 GiB	4.00 GiB	4.25×