Is Qwen3.8-27B or Qwen3.6-34B-80L-Fable-5-Heretic smaller?

Qwen3.8-27B is smaller at Q4_K_M: 15.66 GiB against 19.10 GiB.

Which uses less memory at long context?

Qwen3.8-27B — its KV cache at 32K context is 1.3x smaller, because of how its attention layers cache.

Model comparison

Qwen3.8-27B vs Qwen3.6-34B-80L-Fable-5-Heretic

At Q4_K_M, Qwen3.8-27B is the smaller download — 16,810,714,336 bytes against 20,505,296,736. At long context the gap widens: Qwen3.8-27B's KV cache at 32K is 1.3× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Qwen3.8-27B	Qwen3.6-34B-80L-Fable-5-Heretic
Parameters	27.8B	33.4B
Architecture	qwen35	qwen35
Layers	64	80
Native context	262,144	262,144
Mixture of experts	no	no
Quantizations published	22	25
Smallest quantization	8.39 GiB	12.27 GiB
Q4_K_M	15.66 GiB	19.10 GiB
Licence	apache-2.0	—

KV cache by context

the term that decides long-context viability

Context	Qwen3.8-27B	Qwen3.6-34B-80L-Fable-5-Heretic	Ratio
4,096	0.25 GiB	0.31 GiB	1.25×
8,192	0.50 GiB	0.63 GiB	1.25×
16,384	1.00 GiB	1.25 GiB	1.25×
32,768	2.00 GiB	2.50 GiB	1.25×
65,536	4.00 GiB	5.00 GiB	1.25×
131,072	8.00 GiB	10.00 GiB	1.25×