Is reka-flash-3 or Qwen3.6-27B smaller?

reka-flash-3 is smaller at Q4_K_M: 12.68 GiB against 15.41 GiB.

Which uses less memory at long context?

Qwen3.6-27B — its KV cache at 32K context is 2.1x smaller, because of how its attention layers cache.

Model comparison

reka-flash-3 vs Qwen3.6-27B

At Q4_K_M, reka-flash-3 is the smaller download — 13,610,362,336 bytes against 16,547,397,888. At long context the gap widens: Qwen3.6-27B's KV cache at 32K is 2.1× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	reka-flash-3	Qwen3.6-27B
Parameters	20.9B	27.8B
Architecture	llama	qwen35
Layers	44	64
Native context	32,768	262,144
Mixture of experts	no	no
Quantizations published	38	40
Smallest quantization	6.88 GiB	8.74 GiB
Q4_K_M	12.68 GiB	15.41 GiB
Licence	apache-2.0	apache-2.0

KV cache by context

the term that decides long-context viability

Context	reka-flash-3	Qwen3.6-27B	Ratio
4,096	0.52 GiB	0.25 GiB	2.06×
8,192	1.03 GiB	0.50 GiB	2.06×
16,384	2.06 GiB	1.00 GiB	2.06×
32,768	4.13 GiB	2.00 GiB	2.06×
65,536	8.25 GiB	4.00 GiB	2.06×
131,072	16.50 GiB	8.00 GiB	2.06×