Is granite-4.0-7B-A1B-Creative-v0.1 or Qwen3-8B smaller?

granite-4.0-7B-A1B-Creative-v0.1 is smaller at Q4_K_M: 3.77 GiB against 4.68 GiB.

Which uses less memory at long context?

granite-4.0-7B-A1B-Creative-v0.1 — its KV cache at 32K context is 18.0x smaller, because of how its attention layers cache.

Model comparison

granite-4.0-7B-A1B-Creative-v0.1 vs Qwen3-8B

At Q4_K_M, granite-4.0-7B-A1B-Creative-v0.1 is the smaller download — 4,049,883,328 bytes against 5,027,783,488. At long context the gap widens: granite-4.0-7B-A1B-Creative-v0.1's KV cache at 32K is 18.0× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	granite-4.0-7B-A1B-Creative-v0.1	Qwen3-8B
Parameters	6.7B	8.2B
Architecture	granitehybrid	qwen3
Layers	40	36
Native context	131,072	40,960
Mixture of experts	yes, 62 experts	no
Quantizations published	36	51
Smallest quantization	1.29 GiB	2.12 GiB
Q4_K_M	3.77 GiB	4.68 GiB
Licence	apache-2.0	apache-2.0

KV cache by context

the term that decides long-context viability

Context	granite-4.0-7B-A1B-Creative-v0.1	Qwen3-8B	Ratio
4,096	0.03 GiB	0.56 GiB	17.96×
8,192	0.06 GiB	1.13 GiB	17.98×
16,384	0.13 GiB	2.25 GiB	17.99×
32,768	0.25 GiB	4.50 GiB	18.00×
65,536	0.50 GiB	9.00 GiB	18.00×
131,072	1.00 GiB	18.00 GiB	18.00×