Is bitnet-b1.58-2B-4T or SmolLM2-135M-Instruct smaller?

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

Which uses less memory at long context?

SmolLM2-135M-Instruct — its KV cache at 32K context is 3.3x smaller, because of how its attention layers cache.

Model comparison

bitnet-b1.58-2B-4T vs SmolLM2-135M-Instruct

These two publish different quantization sets; the table below has the exact sizes. At long context the gap widens: SmolLM2-135M-Instruct's KV cache at 32K is 3.3× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	bitnet-b1.58-2B-4T	SmolLM2-135M-Instruct
Parameters	850M	135M
Architecture	—	llama
Layers	30	30
Native context	4,096	8,192
Mixture of experts	no	no
Quantizations published	2	41
Smallest quantization	0.10 GiB	0.08 GiB
Q4_K_M	—	0.10 GiB
Licence	mit	apache-2.0

KV cache by context

the term that decides long-context viability

Context	bitnet-b1.58-2B-4T	SmolLM2-135M-Instruct	Ratio
4,096	0.29 GiB	0.09 GiB	3.33×
8,192	0.59 GiB	0.18 GiB	3.33×
16,384	1.17 GiB	0.35 GiB	3.33×
32,768	2.34 GiB	0.70 GiB	3.33×
65,536	4.69 GiB	1.41 GiB	3.33×
131,072	9.38 GiB	2.81 GiB	3.33×