LES Logo

Lestvica embeddingov za slovenőčino

Comparing embedding models on Slovene NLP tasks

RankModelParametersContext WindowScoreClassificationClusteringRetrieval
πŸ₯‡
google/gemini-embedding-001
-2,04858.1258.9722.6192.77
πŸ₯ˆ
Qwen/Qwen3-Embedding-8BπŸ€—
8B32,76855.7253.3323.5390.30
πŸ₯‰
Snowflake/snowflake-arctic-embed-l-v2.0πŸ€—
0.6B8,19255.6352.3523.9490.59
4
openai/text-embedding-3-large
-8,19255.5855.7124.3786.66
5
BAAI/bge-m3πŸ€—
560M8,19255.0852.4522.6590.15
6
Qwen/Qwen3-Embedding-4BπŸ€—
4B32,76855.0551.9624.4888.73
7
intfloat/multilingual-e5-large-instructπŸ€—
0.6B51254.0852.7623.9385.55
8
Alibaba-NLP/gte-multilingual-baseπŸ€—
0.3B819253.5250.6424.2785.63
9
openai/text-embedding-3-small
-8,19253.3752.0723.8084.24
10
google/text-multilingual-embedding-002
––52.4950.8018.6788.01
11
google/embeddinggemma-300mπŸ€—
0.3B2,04851.5649.3423.0482.31
12
Lajavaness/bilingual-embedding-smallπŸ€—
0.1B51250.9548.4923.0881.26
13
intfloat/multilingual-e5-smallπŸ€—
0.1B51250.4147.6420.8582.72
14
Qwen/Qwen3-Embedding-0.6BπŸ€—
0.6B32,76850.1146.6121.1682.55
15
sentence-transformers/paraphrase-multilingual-mpnet-base-v2πŸ€—
0.3B51246.2751.1521.2866.40
16
sentence-transformers/LaBSEπŸ€—
0.5B51244.7148.8519.8665.43
17
cjvt/GaMS-1BπŸ€—
1B2,04843.3055.2927.8346.78
18
google/text-embedding-005
––41.9139.5215.4170.80
19
intfloat/e5-large-v2πŸ€—
0.3B51239.7939.8515.4964.03
20
Snowflake/snowflake-arctic-embed-lπŸ€—
0.3B8,19237.5538.2513.6260.77
21
NovaSearch/stella/en/400M/v5πŸ€—
0.4B8,19235.4341.6017.9246.78
22
intfloat/e5-small-v2πŸ€—
~33M51233.9335.9514.1551.70
23
cjvt/GaMS-2BπŸ€—
2B-29.8750.1321.3918.08
24
EMBEDDIA/slobertaπŸ€—
~110M51228.2949.8118.5516.52
25
sentence-transformers/all-MiniLM-L6-v2πŸ€—
22.7M51225.6732.6614.6029.76
πŸ₯‡ 1stπŸ₯ˆ 2ndπŸ₯‰ 3rdBold = best in column
Scores are shown as percentages

About the Overall score

The overall score is the mean of the main Classification, Retrieval, and Clustering scores.

πŸ“Read the Blog Post

Thanks to Our Data Providers

CJVTLexpera