Is PaddleOCR-VL-1.6 or Qwen3-0.6B smaller?

PaddleOCR-VL-1.6 is smaller at Q4_K_M: 0.28 GiB against 0.37 GiB.

Which uses less memory at long context?

PaddleOCR-VL-1.6 — its KV cache at 32K context is 6.2x smaller, because of how its attention layers cache.

Model comparison

PaddleOCR-VL-1.6 vs Qwen3-0.6B

At Q4_K_M, PaddleOCR-VL-1.6 is the smaller download — 300,219,456 bytes against 396,705,472. At long context the gap widens: PaddleOCR-VL-1.6's KV cache at 32K is 6.2× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	PaddleOCR-VL-1.6	Qwen3-0.6B
Parameters	959M	752M
Architecture	paddleocr	qwen3
Layers	18	28
Native context	131,072	40,960
Mixture of experts	no	no
Quantizations published	36	44
Smallest quantization	0.15 GiB	0.20 GiB
Q4_K_M	0.28 GiB	0.37 GiB
Licence	apache-2.0	—

KV cache by context

the term that decides long-context viability

Context	PaddleOCR-VL-1.6	Qwen3-0.6B	Ratio
4,096	0.07 GiB	0.44 GiB	6.22×
8,192	0.14 GiB	0.88 GiB	6.22×
16,384	0.28 GiB	1.75 GiB	6.22×
32,768	0.56 GiB	3.50 GiB	6.22×
65,536	1.13 GiB	7.00 GiB	6.22×
131,072	2.25 GiB	14.00 GiB	6.22×