Is granite-4.0-h-1b or Llama-3.2-1B-Instruct smaller?

Llama-3.2-1B-Instruct is smaller at Q4_K_M: 0.75 GiB against 0.84 GiB.

Which uses less memory at long context?

granite-4.0-h-1b — its KV cache at 32K context is 4.0x smaller, because of how its attention layers cache.

Model comparison

granite-4.0-h-1b vs Llama-3.2-1B-Instruct

At Q4_K_M, Llama-3.2-1B-Instruct is the smaller download — 807,690,688 bytes against 901,162,208. At long context the gap widens: granite-4.0-h-1b's KV cache at 32K is 4.0× 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-1b	Llama-3.2-1B-Instruct
Parameters	1.5B	1.2B
Architecture	granitehybrid	llama
Layers	40	16
Native context	131,072	131,072
Mixture of experts	no	no
Quantizations published	35	39
Smallest quantization	0.36 GiB	0.39 GiB
Q4_K_M	0.84 GiB	0.75 GiB
Licence	apache-2.0	—

KV cache by context

the term that decides long-context viability

Context	granite-4.0-h-1b	Llama-3.2-1B-Instruct	Ratio
4,096	0.03 GiB	0.13 GiB	3.99×
8,192	0.06 GiB	0.25 GiB	4.00×
16,384	0.13 GiB	0.50 GiB	4.00×
32,768	0.25 GiB	1.00 GiB	4.00×
65,536	0.50 GiB	2.00 GiB	4.00×
131,072	1.00 GiB	4.00 GiB	4.00×