Model comparison

snowflake-arctic-embed-l-v2.0 vs jina-embeddings-v5-text-small

At Q4_K_M, jina-embeddings-v5-text-small is the smaller download — 396,705,152 bytes against 437,779,488. At long context the gap widens: snowflake-arctic-embed-l-v2.0's KV cache at 32K is 1.2× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

snowflake-arctic-embed-l-v2.0jina-embeddings-v5-text-small
Parameters568M596M
Architecturebertqwen3
Layers2428
Native context8,19432,768
Mixture of expertsnono
Quantizations published1556
Smallest quantization0.39 GiB0.19 GiB
Q4_K_M0.41 GiB0.37 GiB
Licence—cc-by-nc-4.0

KV cache by context

the term that decides long-context viability
Contextsnowflake-arctic-embed-l-v2.0jina-embeddings-v5-text-smallRatio
4,0960.38 GiB0.44 GiB1.17×
8,1920.75 GiB0.88 GiB1.17×
16,3841.50 GiB1.75 GiB1.17×
32,7683.00 GiB3.50 GiB1.17×
65,5366.00 GiB7.00 GiB1.17×
131,07212.00 GiB14.00 GiB1.17×