Is Qwen3.5-REAP-262B-A17B or DeepSeek-V4-Flash smaller?

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

Which uses less memory at long context?

DeepSeek-V4-Flash — its KV cache at 32K context is 14.9x smaller, because of how its attention layers cache.

Model comparison

Qwen3.5-REAP-262B-A17B vs DeepSeek-V4-Flash

These two publish different quantization sets; the table below has the exact sizes. At long context the gap widens: DeepSeek-V4-Flash's KV cache at 32K is 14.9× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Qwen3.5-REAP-262B-A17B	DeepSeek-V4-Flash
Parameters	262B	291B
Architecture	qwen35moe	deepseek4
Layers	60	43
Native context	262,144	1,048,576
Mixture of experts	yes, 333 experts	yes, 256 experts
Quantizations published	10	12
Smallest quantization	64.50 GiB	76.87 GiB
Q4_K_M	147.71 GiB	—
Licence	apache-2.0	mit

KV cache by context

the term that decides long-context viability

Context	Qwen3.5-REAP-262B-A17B	DeepSeek-V4-Flash	Ratio
4,096	0.12 GiB	0.06 GiB	1.86×
8,192	0.23 GiB	0.06 GiB	3.72×
16,384	0.47 GiB	0.06 GiB	7.44×
32,768	0.94 GiB	0.06 GiB	14.89×
65,536	1.88 GiB	0.06 GiB	29.77×
131,072	3.75 GiB	0.06 GiB	59.54×