Is functiongemma-270m-it or Qwen2.5-0.5B-Instruct smaller?

functiongemma-270m-it is smaller at Q4_K_M: 0.24 GiB against 0.37 GiB.

Which uses less memory at long context?

functiongemma-270m-it — its KV cache at 32K context is 3.5x smaller, because of how its attention layers cache.

Model comparison

functiongemma-270m-it vs Qwen2.5-0.5B-Instruct

At Q4_K_M, functiongemma-270m-it is the smaller download — 253,127,392 bytes against 397,807,936. At long context the gap widens: functiongemma-270m-it's KV cache at 32K is 3.5× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	functiongemma-270m-it	Qwen2.5-0.5B-Instruct
Parameters	268M	494M
Architecture	gemma3	qwen2
Layers	18	24
Native context	32,768	32,768
Mixture of experts	no	no
Quantizations published	45	32
Smallest quantization	0.17 GiB	0.31 GiB
Q4_K_M	0.24 GiB	0.37 GiB
Licence	gemma	apache-2.0

KV cache by context

the term that decides long-context viability

Context	functiongemma-270m-it	Qwen2.5-0.5B-Instruct	Ratio
4,096	0.03 GiB	0.05 GiB	1.78×
8,192	0.04 GiB	0.09 GiB	2.46×
16,384	0.06 GiB	0.19 GiB	3.05×
32,768	0.11 GiB	0.38 GiB	3.46×
65,536	0.20 GiB	0.75 GiB	3.71×
131,072	0.39 GiB	1.50 GiB	3.85×