Is OYM-Qimi-122B-A10B-K2.6 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 11.9x smaller, because of how its attention layers cache.

Model comparison

OYM-Qimi-122B-A10B-K2.6 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 11.9× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	OYM-Qimi-122B-A10B-K2.6	DeepSeek-V4-Flash
Parameters	125B	291B
Architecture	qwen35moe	deepseek4
Layers	48	43
Native context	262,144	1,048,576
Mixture of experts	yes, 256 experts	yes, 256 experts
Quantizations published	22	12
Smallest quantization	42.67 GiB	76.87 GiB
Q4_K_M	70.64 GiB	—
Licence	other	mit

KV cache by context

the term that decides long-context viability

Context	OYM-Qimi-122B-A10B-K2.6	DeepSeek-V4-Flash	Ratio
4,096	0.09 GiB	0.06 GiB	1.49×
8,192	0.19 GiB	0.06 GiB	2.98×
16,384	0.38 GiB	0.06 GiB	5.95×
32,768	0.75 GiB	0.06 GiB	11.91×
65,536	1.50 GiB	0.06 GiB	23.82×
131,072	3.00 GiB	0.06 GiB	47.63×