Is gte-large or jina-embeddings-v5-text-nano smaller?

jina-embeddings-v5-text-nano is smaller at Q4_K_M: 0.15 GiB against 0.20 GiB.

Which uses less memory at long context?

jina-embeddings-v5-text-nano — its KV cache at 32K context is 2.7x smaller, because of how its attention layers cache.

Model comparison

gte-large vs jina-embeddings-v5-text-nano

At Q4_K_M, jina-embeddings-v5-text-nano is the smaller download — 156,966,976 bytes against 215,891,488. At long context the gap widens: jina-embeddings-v5-text-nano's KV cache at 32K is 2.7× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	gte-large	jina-embeddings-v5-text-nano
Parameters	335M	212M
Architecture	bert	eurobert
Layers	24	12
Native context	512	8,192
Mixture of experts	no	no
Quantizations published	12	42
Smallest quantization	0.13 GiB	0.09 GiB
Q4_K_M	0.20 GiB	0.15 GiB
Licence	mit	cc-by-nc-4.0

KV cache by context

the term that decides long-context viability

Context	gte-large	jina-embeddings-v5-text-nano	Ratio
4,096	0.38 GiB	0.14 GiB	2.67×
8,192	0.75 GiB	0.28 GiB	2.67×
16,384	1.50 GiB	0.56 GiB	2.67×
32,768	3.00 GiB	1.13 GiB	2.67×
65,536	6.00 GiB	2.25 GiB	2.67×
131,072	12.00 GiB	4.50 GiB	2.67×