Bge Base En v1.5

Sep 2023
Bge Base En v1.5
Compare in
on
$0.005per 1M tokens
54.8MTEB English v2
Sep 2023released
768Dimensions
1K tokensMax input
enLanguages

Rankingconfiguration unstated?

Full ranking →
1Ja Embed Sts85.62F2LLM v2 8B85.43F2LLM v2 14B85.14Harrier OSS v1 0.6B84.85F2LLM v2 4B84.7
···
100Bge Large En v1.562.2101Bge Base En v1.561.6102MedEmbed Small v0.161.6

Strong at

B
Embeddings & Retrieval#63 of 200

Alternatives?

ModelScoreCostPriceWhat changes
CheaperEmbeddingGemma 300MunstatedOpen+37%-60%$0.002Languages: not verified+1NOpenJa Embed Sts+39%——Memory (4-bit): 0.1 GB → 0.2 GBsBetterAll Mpnet Base v2unstatedOpen+9%same$0.005Max input: 1K tokens → 0K tokenssBetterAll MiniLM L6 v2unstatedOpen+9%same$0.005Max input: 1K tokens → 0K tokenssBetterAll MiniLM L12 v2unstatedOpen+9%same$0.005Max input: 1K tokens → 0K tokens

Run it yourself

estimate
Fits on✓ 8 GB (phone, laptop) · full quality✓ 16 GB · full quality✓ 24 GB GPU · full quality✓ 48 GB GPU · full quality✓ 80 GB GPU · full quality
LicenceMITfree for commercial use

Benchmarks

in Embeddings & Retrieval
BEIR · Mean over tasks · English53.2#64
BEIR · Mean over task types · English53.2#64
BEIR · Retrieval · English53.2#64
ChemTEB · Mean over tasks · multilingual (13 languages)70.5#22
ChemTEB · Mean over task types · multilingual (13 languages)51.7#15
ChemTEB · Classification · English81.1#27
ChemTEB · Retrieval · English73.7#15
ChemTEB · Pair Classification · multilingual (13 languages)56.7#23
ChemTEB · Clustering · English46.6#19
ChemTEB · Bitext Mining · English0.2#27
MTEB Persian v2 · Reranking · Persian38.5#80
JMTEB v2 · Clustering · Japanese18.2#104
JMTEB v2 · STS · Japanese61.6#101
LongEmbed · Mean over tasks · English33.9#110
LongEmbed · Mean over task types · English33.9#110
LongEmbed · Retrieval · English33.9#110
MTEB German v1 · Mean over tasks · German40.5#54
MTEB German v1 · Mean over task types · German39.4#54
MTEB German v1 · Clustering · German25.2#79
MTEB German v1 · Retrieval · German38.9#51
MTEB German v1 · Classification · German48.5#105
MTEB German v1 · STS · German49.1#88
MTEB German v1 · Pair Classification · German48.9#135
MTEB German v1 · Reranking · German25.8#81
MTEB Europe v1 · Mean over tasks · multilingual (29 languages)46.9#43
MTEB Europe v1 · Mean over task types · multilingual (29 languages)41.3#43
MTEB Europe v1 · Retrieval · multilingual (25 languages)40.8#74
MTEB Europe v1 · STS · multilingual (8 languages)66.6#48
MTEB Europe v1 · Bitext Mining · multilingual (28 languages)34.1#130
MTEB Europe v1 · Classification · multilingual (23 languages)50.8#80
MTEB Europe v1 · Clustering · multilingual (29 languages)37.1#84
MTEB Europe v1 · Instruction Reranking · English-4#195
MTEB Europe v1 · Pair Classification · multilingual (11 languages)83.1#85
MTEB Europe v1 · Reranking · multilingual (12 languages)50.4#106
MTEB Europe v1 · Multilabel Classification · multilingual (23 languages)12.6#95
MTEB French v1 · Mean over tasks · multilingual (2 languages)49.2#52
MTEB French v1 · Mean over task types · multilingual (2 languages)50.9#50
MTEB French v1 · STS · French70.2#60
MTEB French v1 · Retrieval · multilingual (2 languages)36.7#73
MTEB French v1 · Reranking · French67.6#74
MTEB French v1 · Clustering · French42.1#47
MTEB French v1 · Classification · French53.4#80
MTEB French v1 · Summarization · French30.7#47
MTEB French v1 · Pair Classification · French55.6#132
MTEB Indic v1 · Mean over tasks · multilingual (25 languages)33.9#77
MTEB Indic v1 · Mean over task types · multilingual (25 languages)25.9#64
MTEB Indic v1 · Classification · multilingual (8 languages)43.3#92
MTEB Indic v1 · Retrieval · multilingual (14 languages)16.7#85
MTEB Indic v1 · Bitext Mining · multilingual (20 languages)3.4#115
MTEB Indic v1 · Reranking · multilingual (2 languages)53.6#119
MTEB Indic v1 · Pair Classification · Hindi54.1#123
MTEB Indic v1 · STS · multilingual (13 languages)7.4#113
MTEB Indic v1 · Clustering · multilingual (21 languages)3.1#159
MTEB Korean v2 · STS · Korean46.9#51
MTEB Korean v2 · Reranking · Korean19.9#72
MTEB Korean v2 · Classification · Korean21.3#110
MTEB Dutch v1 · Mean over tasks · Dutch41#76
MTEB Dutch v1 · Mean over task types · Dutch48.3#80
MTEB Dutch v1 · Classification · Dutch46.7#91
MTEB Dutch v1 · Clustering · Dutch23.7#78
MTEB Dutch v1 · Retrieval · Dutch40.5#70
MTEB Dutch v1 · STS · Dutch58.9#91
MTEB Dutch v1 · Multilabel Classification · Dutch30.9#82
MTEB Dutch v1 · Pair Classification · Dutch58.4#122
MTEB Dutch v1 · Reranking · Dutch78.8#127
MTEB Polish v1 · Pair Classification · Polish62.6#112
MTEB Polish v1 · Classification · Polish42.3#82
MTEB Polish v1 · STS · Polish56.9#67
MTEB Portuguese v1 · Multilabel Classification · Portuguese18.7#124
MTEB Law v1 · Mean over tasks · multilingual (3 languages)40.2#83
MTEB Law v1 · Mean over task types · multilingual (3 languages)40.2#83
MTEB Law v1 · Retrieval · multilingual (3 languages)40.2#83
MTEB Medical v1 · Clustering · English35.9#104
MTEB Medical v1 · Reranking · Mandarin Chinese22.1#124
FollowIR · Mean over tasks · English-4#195
FollowIR · Mean over task types · English-4#195
FollowIR · Instruction Reranking · English-4#195
MTEB Spanish v1 · Pair Classification · Spanish59.3#113
MTEB Spanish v1 · STS · multilingual (2 languages)62.9#54
MTEB Spanish v1 · Clustering · Spanish40.6#80
MTEB Spanish v1 · Reranking · Spanish38.4#78
MTEB Thai v1 · Clustering · Thai3.2#179
MTEB Thai v1 · Pair Classification · Thai51.8#126
MTEB Scandinavian v1 · Clustering · multilingual (2 languages)33.3#94
MTEB Scandinavian v1 · Classification · multilingual (6 languages)48.2#109
MTEB Scandinavian v1 · Bitext Mining · multilingual (3 languages)59.1#143
MTEB Russian v1.1 · Mean over tasks · Russian29.6#66
MTEB Russian v1.1 · Mean over task types · Russian30.6#66
MTEB Russian v1.1 · Classification · Russian31.9#105
MTEB Russian v1.1 · STS · Russian41.5#101
MTEB Russian v1.1 · Reranking · Russian33.5#87
MTEB Russian v1.1 · Clustering · Russian17.6#113
MTEB Russian v1.1 · Pair Classification · Russian47.1#166
MTEB Russian v1.1 · Retrieval · Russian17#57
MTEB Russian v1.1 · Multilabel Classification · Russian25.8#127
RTEB Finnish beta · Mean over tasks · English54.7#55
RTEB French beta · Mean over tasks · multilingual (2 languages)23.4#116
RTEB German beta · Mean over tasks · German16#143
RTEB Health beta · Mean over tasks · multilingual (4 languages)43.6#61
RTEB Law beta · Mean over tasks · multilingual (4 languages)32.5#123
MTEB multilingual v2 · Mean over tasks · multilingual (1037 languages)44.5#69
MTEB multilingual v2 · Mean over task types · multilingual (1037 languages)38.1#68
MTEB multilingual v2 · Classification · multilingual (119 languages)48#104
MTEB multilingual v2 · Clustering · multilingual (208 languages)41.4#77
MTEB multilingual v2 · STS · multilingual (35 languages)60#77
MTEB multilingual v2 · Pair Classification · multilingual (24 languages)71.7#125
MTEB multilingual v2 · Retrieval · multilingual (120 languages)38.6#103
MTEB multilingual v2 · Bitext Mining · multilingual (1026 languages)23.7#126
MTEB multilingual v2 · Instruction Reranking · English-4#195
MTEB multilingual v2 · Multilabel Classification · multilingual (25 languages)16.4#107
MTEB multilingual v2 · Reranking · multilingual (22 languages)47.6#99

FAQ

How much does Bge Base En v1.5 cost?›

$0.005 per 1M tokens.

Is Bge Base En v1.5 free to use?›

There is no free API; the cheapest paid price is $0.005 per 1M input tokens. The weights are open, so you can run it yourself.

How good is Bge Base En v1.5, and what is it best at?›

#63 of 200 in Embeddings.

Is Bge Base En v1.5 better than Bge Large En v1.5?›

On JMTEB v2 · STS · Japanese, Bge Base En v1.5 scores 61.6 vs 62.2 — about the same. It is 50% cheaper.

How can I use Bge Base En v1.5?›

Through the BAAI API.

When was Bge Base En v1.5 released? Is there a newer version?›

Released 2023-09-11. It is the current version in its line.

Is Bge Base En v1.5 open source?›

Yes, the weights are open (MIT licence).

Can I run Bge Base En v1.5 locally?›

The weights are open and can be self-hosted; hardware needs are not confirmed yet.

What is Bge Base En v1.5?›

Bge Base En v1.5 is an open-weights embeddings & retrieval model from BAAI, released Sep 2023.

What are the best alternatives to Bge Base En v1.5?›

EmbeddingGemma 300M (cheaper), Ja Embed Sts (open weights).