Is HarmonicHarlequin_v5-20B or gemma-4-12B-it-qat-q4_0-unquantized smaller?

Both publish quantizations; compare the table for the exact sizes.

Which uses less memory at long context?

gemma-4-12B-it-qat-q4_0-unquantized — its KV cache at 32K context is 26.3x smaller, because of how its attention layers cache.

Model comparison

HarmonicHarlequin_v5-20B vs gemma-4-12B-it-qat-q4_0-unquantized

These two publish different quantization sets; the table below has the exact sizes. At long context the gap widens: gemma-4-12B-it-qat-q4_0-unquantized's KV cache at 32K is 26.3× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	HarmonicHarlequin_v5-20B	gemma-4-12B-it-qat-q4_0-unquantized
Parameters	33.3B	12.0B
Architecture	llama	gemma4
Layers	104	48
Native context	4,096	262,144
Mixture of experts	no	no
Quantizations published	35	2
Smallest quantization	6.77 GiB	6.50 GiB
Q4_K_M	18.71 GiB	—
Licence	—	apache-2.0

KV cache by context

the term that decides long-context viability

Context	HarmonicHarlequin_v5-20B	gemma-4-12B-it-qat-q4_0-unquantized	Ratio
4,096	8.13 GiB	0.72 GiB	11.30×
8,192	16.25 GiB	0.97 GiB	16.77×
16,384	32.50 GiB	1.47 GiB	22.13×
32,768	65.00 GiB	2.47 GiB	26.33×
65,536	130.00 GiB	4.47 GiB	29.09×
131,072	260.00 GiB	8.47 GiB	30.70×