Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

MTEB

Benchmarks

Task NameDataset NameSOTA ResultTrend
Text EmbeddingMTEB English v2
Mean Score74.6
107
Sentence Embedding EvaluationMTEB (test)
Classification Score90.37
55
Text EmbeddingMTEB
Classification Score72.71
50
Multilingual RetrievalMTEB Multilingual v2
nDCG@1070.9
40
RetrievalMTEB eng v2
nDCG@1069.4
31
Text EmbeddingMTEB Code v1
Average Performance70
30
Multilingual Text EmbeddingMTEB Multilingual
Mean Score (Task)72.3
29
Information RetrievalMTEB v2
NDCG@1045.9
28
Code RetrievalMTEB Code Retrieval Average (val)
nDCG@574
24
Text EmbeddingMTEB Turkish (test)
Overall MTEB Score65.42
23
ClusteringMTEB Clustering
Bior Score33.13
23
Text Embedding EvaluationMTEB eng v2 (test)
Average Score67.3
22
Code RetrievalMTEB Code
nDCG@1080.07
21
Semantic Textual Similarity (STS)MTEB English 2023 (test)
BIO89.37
19
ClusteringMTEB Clustering v1 (test)
TNG58.14
18
Text EmbeddingMTEB v2
Clustering Score42.5
17
Text EmbeddingMTEB Multilingual V2 (test)
Mean Score (TaskType)62.5
16
Embedding EvaluationMTEB Corrupt
Classification Score40
15
Embedding EvaluationMTEB Clean
Classification Score51.6
15
Text EmbeddingMTEB (test)
Average Score72.97
14
Text Embedding EvaluationMTEB [hye]
Score78.04
13
Text ClassificationMTEB classification (test)
Emotion Score67
12
Text EmbeddingMTEB
Average (Multi-Language/Domain) Performance72.32
12
Code RetrievalMTEB Code (test)
Apps Score98.1
12
Sentence EmbeddingMTEB Clustering standard (test)
AskU. Score63.48
12
Showing 25 of 57 rows