Is LFM2-1.2B-RAG or Qwen3-4B smaller?

LFM2-1.2B-RAG is smaller at Q4_K_M: 0.68 GiB against 2.33 GiB.

Which uses less memory at long context?

LFM2-1.2B-RAG — its KV cache at 32K context is 12.0x smaller, because of how its attention layers cache.

Model comparison

LFM2-1.2B-RAG vs Qwen3-4B

At Q4_K_M, LFM2-1.2B-RAG is the smaller download — 730,894,048 bytes against 2,497,280,256. At long context the gap widens: LFM2-1.2B-RAG's KV cache at 32K is 12.0× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	LFM2-1.2B-RAG	Qwen3-4B
Parameters	1.2B	4.0B
Architecture	lfm2	qwen3
Layers	16	36
Native context	128,000	40,960
Mixture of experts	no	no
Quantizations published	29	29
Smallest quantization	0.40 GiB	1.01 GiB
Q4_K_M	0.68 GiB	2.33 GiB
Licence	other	—

KV cache by context

the term that decides long-context viability

Context	LFM2-1.2B-RAG	Qwen3-4B	Ratio
4,096	0.05 GiB	0.56 GiB	12.00×
8,192	0.09 GiB	1.13 GiB	12.00×
16,384	0.19 GiB	2.25 GiB	12.00×
32,768	0.38 GiB	4.50 GiB	12.00×
65,536	0.75 GiB	9.00 GiB	12.00×
131,072	1.50 GiB	18.00 GiB	12.00×