Is Fourier-Qwen2.5-VL-3B-0.67 or Qwen3-4B smaller?

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

Which uses less memory at long context?

Fourier-Qwen2.5-VL-3B-0.67 — its KV cache at 32K context is 4.0x smaller, because of how its attention layers cache.

Model comparison

Fourier-Qwen2.5-VL-3B-0.67 vs Qwen3-4B

These two publish different quantization sets; the table below has the exact sizes. At long context the gap widens: Fourier-Qwen2.5-VL-3B-0.67's KV cache at 32K is 4.0× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Fourier-Qwen2.5-VL-3B-0.67	Qwen3-4B
Parameters	3.8B	4.0B
Architecture	qwen2vl	qwen3
Layers	36	36
Native context	128,000	40,960
Mixture of experts	no	no
Quantizations published	24	29
Smallest quantization	0.74 GiB	1.01 GiB
Q4_K_M	—	2.33 GiB
Licence	apache-2.0	—

KV cache by context

the term that decides long-context viability

Context	Fourier-Qwen2.5-VL-3B-0.67	Qwen3-4B	Ratio
4,096	0.14 GiB	0.56 GiB	4.00×
8,192	0.28 GiB	1.13 GiB	4.00×
16,384	0.56 GiB	2.25 GiB	4.00×
32,768	1.13 GiB	4.50 GiB	4.00×
65,536	2.25 GiB	9.00 GiB	4.00×
131,072	4.50 GiB	18.00 GiB	4.00×