Is granite-embedding-107m-multilingual or Qwen3-0.6B smaller?

granite-embedding-107m-multilingual is smaller at Q4_K_M: 0.11 GiB against 0.37 GiB.

Which uses less memory at long context?

granite-embedding-107m-multilingual — its KV cache at 32K context is 12.4x smaller, because of how its attention layers cache.

Model comparison

granite-embedding-107m-multilingual vs Qwen3-0.6B

At Q4_K_M, granite-embedding-107m-multilingual is the smaller download — 117,011,136 bytes against 396,705,472. At long context the gap widens: granite-embedding-107m-multilingual's KV cache at 32K is 12.4× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	granite-embedding-107m-multilingual	Qwen3-0.6B
Parameters	107M	752M
Architecture	bert	qwen3
Layers	6	28
Native context	514	40,960
Mixture of experts	no	no
Quantizations published	19	44
Smallest quantization	0.11 GiB	0.20 GiB
Q4_K_M	0.11 GiB	0.37 GiB
Licence	apache-2.0	—

KV cache by context

the term that decides long-context viability

Context	granite-embedding-107m-multilingual	Qwen3-0.6B	Ratio
4,096	0.04 GiB	0.44 GiB	12.44×
8,192	0.07 GiB	0.88 GiB	12.44×
16,384	0.14 GiB	1.75 GiB	12.44×
32,768	0.28 GiB	3.50 GiB	12.44×
65,536	0.56 GiB	7.00 GiB	12.44×
131,072	1.13 GiB	14.00 GiB	12.44×