Is Qwen3.6-27B or Delphi-25B-SimpleRL-Math smaller?

Delphi-25B-SimpleRL-Math is smaller at Q4_K_M: 14.12 GiB against 15.41 GiB.

Which uses less memory at long context?

Qwen3.6-27B — its KV cache at 32K context is 16.7x smaller, because of how its attention layers cache.

Model comparison

Qwen3.6-27B vs Delphi-25B-SimpleRL-Math

At Q4_K_M, Delphi-25B-SimpleRL-Math is the smaller download — 15,160,698,624 bytes against 16,547,397,888. At long context the gap widens: Qwen3.6-27B's KV cache at 32K is 16.7× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Qwen3.6-27B	Delphi-25B-SimpleRL-Math
Parameters	27.8B	25.0B
Architecture	qwen35	qwen3
Layers	64	51
Native context	262,144	16,384
Mixture of experts	no	no
Quantizations published	40	34
Smallest quantization	8.74 GiB	5.31 GiB
Q4_K_M	15.41 GiB	14.12 GiB
Licence	apache-2.0	—

KV cache by context

the term that decides long-context viability

Context	Qwen3.6-27B	Delphi-25B-SimpleRL-Math	Ratio
4,096	0.25 GiB	4.18 GiB	16.72×
8,192	0.50 GiB	8.37 GiB	16.73×
16,384	1.00 GiB	16.73 GiB	16.73×
32,768	2.00 GiB	33.47 GiB	16.73×
65,536	4.00 GiB	66.94 GiB	16.73×
131,072	8.00 GiB	133.88 GiB	16.73×