Is Kimi-K2.6 or Inkling smaller?

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

Which uses less memory at long context?

Kimi-K2.6 — its KV cache at 32K context is 3.8x smaller, because of how its attention layers cache.

Model comparison

Kimi-K2.6 vs Inkling

These two publish different quantization sets; the table below has the exact sizes. At long context the gap widens: Kimi-K2.6's KV cache at 32K is 3.8× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Kimi-K2.6	Inkling
Parameters	1059B	952B
Architecture	deepseek2	inkling
Layers	61	66
Native context	262,144	—
Mixture of experts	yes, 384 experts	yes, 256 experts
Quantizations published	16	7
Smallest quantization	193.16 GiB	210.69 GiB
Q4_K_M	—	—
Licence	other	apache-2.0

KV cache by context

the term that decides long-context viability

Context	Kimi-K2.6	Inkling	Ratio
4,096	0.27 GiB	1.03 GiB	3.85×
8,192	0.54 GiB	2.06 GiB	3.85×
16,384	1.07 GiB	4.13 GiB	3.85×
32,768	2.14 GiB	8.25 GiB	3.85×
65,536	4.29 GiB	16.50 GiB	3.85×
131,072	8.58 GiB	33.00 GiB	3.85×