Is HunyuanOCR or Qwen3-1.7B smaller?

HunyuanOCR is smaller at Q4_K_M: 0.33 GiB against 1.03 GiB.

Which uses less memory at long context?

HunyuanOCR — its KV cache at 32K context is 4.1x smaller, because of how its attention layers cache.

Model comparison

HunyuanOCR vs Qwen3-1.7B

At Q4_K_M, HunyuanOCR is the smaller download — 354,966,656 bytes against 1,107,409,472. At long context the gap widens: HunyuanOCR's KV cache at 32K is 4.1× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	HunyuanOCR	Qwen3-1.7B
Parameters	1.1B	2.0B
Architecture	hunyuan_vl	qwen3
Layers	24	28
Native context	131,072	40,960
Mixture of experts	yes, 1 experts	no
Quantizations published	25	48
Smallest quantization	0.24 GiB	0.50 GiB
Q4_K_M	0.33 GiB	1.03 GiB
Licence	—	—

KV cache by context

the term that decides long-context viability

Context	HunyuanOCR	Qwen3-1.7B	Ratio
4,096	0.11 GiB	0.44 GiB	4.15×
8,192	0.21 GiB	0.88 GiB	4.15×
16,384	0.42 GiB	1.75 GiB	4.15×
32,768	0.84 GiB	3.50 GiB	4.15×
65,536	1.69 GiB	7.00 GiB	4.15×
131,072	3.38 GiB	14.00 GiB	4.15×