Is granite-4.0-h-350m or Qwen3-0.6B smaller?

granite-4.0-h-350m is smaller at Q4_K_M: 0.21 GiB against 0.37 GiB.

Which uses less memory at long context?

granite-4.0-h-350m — its KV cache at 32K context is 28.0x smaller, because of how its attention layers cache.

Model comparison

granite-4.0-h-350m vs Qwen3-0.6B

At Q4_K_M, granite-4.0-h-350m is the smaller download — 222,662,560 bytes against 396,705,472. At long context the gap widens: granite-4.0-h-350m's KV cache at 32K is 28.0× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	granite-4.0-h-350m	Qwen3-0.6B
Parameters	340M	752M
Architecture	granitehybrid	qwen3
Layers	32	28
Native context	32,768	40,960
Mixture of experts	no	no
Quantizations published	29	44
Smallest quantization	0.15 GiB	0.20 GiB
Q4_K_M	0.21 GiB	0.37 GiB
Licence	apache-2.0	—

KV cache by context

the term that decides long-context viability

Context	granite-4.0-h-350m	Qwen3-0.6B	Ratio
4,096	0.02 GiB	0.44 GiB	27.95×
8,192	0.03 GiB	0.88 GiB	27.98×
16,384	0.06 GiB	1.75 GiB	27.99×
32,768	0.13 GiB	3.50 GiB	27.99×
65,536	0.25 GiB	7.00 GiB	28.00×
131,072	0.50 GiB	14.00 GiB	28.00×