Snowflake Arctic Embed M Long

Apr 2024
Snowflake Arctic Embed M Long
Compare in
on
self-host onlyprice
57MTEB English v2
Apr 2024released
8K tokensMax input

Rankingconfiguration unstated?

Full ranking →
1F2LLM v2 14B68.92Gte Qwen2 7B Instruct68.53F2LLM v2 8B68.14F2LLM v2 4B67.85Gte Qwen1.5 7B Instruct66
···
94All Mpnet Base v230.195Snowflake Arctic Embed M Long3096Mmlw Roberta Large29.7

Strong at

A
Embeddings & Retrieval#44 of 200

Alternatives?

ModelScoreCostPriceWhat changes
iBetterE5 Base v2unstatedOpen+12%—$0.005Max input: 8K tokens → 1K tokensOpenGte Qwen2 7B Instructunstated×2.3——Memory (4-bit): 0.1 GB → 4.2 GBtBetterGte BaseunstatedOpen+12%—$0.005Max input: 8K tokens → 1K tokensiBetterMultilingual E5 Large InstructunstatedOpen+94%—$0.01Max input: 8K tokens → 1K tokens+1tBetterGte LargeunstatedOpen+15%—$0.01Max input: 8K tokens → 1K tokens+1

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
LicenceApache-2.0free for commercial use

Benchmarks

in Embeddings & Retrieval
BEIR · Mean over tasks · English54.8#50
BEIR · Mean over task types · English54.8#50
BEIR · Retrieval · English54.8#50
JMTEB v2 · Clustering · Japanese20.6#95
JMTEB v2 · STS · Japanese55.5#128
LongEmbed · Mean over tasks · English45.7#51
LongEmbed · Mean over task types · English45.7#51
LongEmbed · Retrieval · English45.7#51
MTEB German v1 · Classification · German48.7#106
MTEB Europe v1 · Classification · multilingual (23 languages)48.1#108
MTEB Europe v1 · Instruction Reranking · English-1.9#123
MTEB Europe v1 · Multilabel Classification · multilingual (23 languages)12.4#100
MTEB Europe v1 · Pair Classification · multilingual (11 languages)80.5#124
MTEB Europe v1 · Clustering · multilingual (29 languages)35#115
MTEB Europe v1 · Retrieval · multilingual (25 languages)42.4#76
MTEB Europe v1 · Reranking · multilingual (12 languages)48.5#127
MTEB Europe v1 · Bitext Mining · multilingual (28 languages)33.1#134
MTEB French v1 · Pair Classification · French54.5#154
MTEB Indic v1 · Bitext Mining · multilingual (20 languages)3#125
MTEB Indic v1 · STS · multilingual (13 languages)10#104
MTEB Indic v1 · Reranking · multilingual (2 languages)46.6#140
MTEB Indic v1 · Retrieval · multilingual (14 languages)14#104
MTEB Indic v1 · Pair Classification · Hindi52.6#152
MTEB Indic v1 · Clustering · multilingual (21 languages)3.3#145
MTEB Dutch v1 · Reranking · Dutch75.9#150
MTEB Portuguese v1 · Multilabel Classification · Portuguese19.6#97
MTEB Law v1 · Mean over tasks · multilingual (3 languages)37.5#109
MTEB Law v1 · Mean over task types · multilingual (3 languages)37.5#109
MTEB Law v1 · Retrieval · multilingual (3 languages)37.5#109
MTEB Medical v1 · Mean over tasks · multilingual (10 languages)38.4#78
MTEB Medical v1 · Mean over task types · multilingual (10 languages)30#95
MTEB Medical v1 · Retrieval · multilingual (10 languages)42.1#75
MTEB Medical v1 · Clustering · English33.6#151
MTEB Medical v1 · Reranking · Mandarin Chinese14.4#171
FollowIR · Mean over tasks · English-1.9#123
FollowIR · Mean over task types · English-1.9#123
FollowIR · Instruction Reranking · English-1.9#123
MTEB Spanish v1 · Pair Classification · Spanish55.5#156
MTEB Thai v1 · Pair Classification · Thai50.9#178
MTEB Thai v1 · Clustering · Thai3.3#167
MTEB Scandinavian v1 · Classification · multilingual (6 languages)48.1#117
MTEB Scandinavian v1 · Clustering · multilingual (2 languages)31.4#120
MTEB Scandinavian v1 · Bitext Mining · multilingual (3 languages)58.2#151
MTEB Russian v1.1 · Pair Classification · Russian46.3#176
RTEB French beta · Mean over tasks · multilingual (2 languages)29#96
RTEB German beta · Mean over tasks · German9.1#191
RTEB Law beta · Mean over tasks · multilingual (4 languages)29.4#157
MTEB multilingual v2 · Retrieval · multilingual (120 languages)39.3#101
MTEB multilingual v2 · Multilabel Classification · multilingual (25 languages)16.3#108
MTEB multilingual v2 · Bitext Mining · multilingual (1026 languages)22.8#138
MTEB multilingual v2 · Instruction Reranking · English-1.9#123
MTEB multilingual v2 · Reranking · multilingual (22 languages)41.7#131
MTEB multilingual v2 · Pair Classification · multilingual (24 languages)71.7#127

FAQ

How much does Snowflake Arctic Embed M Long cost?›

There is no per-unit API price: self-host only.

Is Snowflake Arctic Embed M Long free to use?›

There is no free API. The weights are open, so you can run it yourself.

How good is Snowflake Arctic Embed M Long, and what is it best at?›

#44 of 200 in Embeddings. Strongest at: Embeddings & Retrieval.

Is Snowflake Arctic Embed M Long better than All Mpnet Base v2?›

On MTEB Medical v1 · Mean over task types · multilingual (10 languages), Snowflake Arctic Embed M Long scores 30 vs 30.1 — about the same.

When was Snowflake Arctic Embed M Long released? Is there a newer version?›

Released 2024-04-12. It is the current version in its line.

Is Snowflake Arctic Embed M Long open source?›

Yes, the weights are open (Apache-2.0 licence).

Can I run Snowflake Arctic Embed M Long locally?›

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

What is Snowflake Arctic Embed M Long?›

Snowflake Arctic Embed M Long is an open-weights embeddings & retrieval model from Snowflake, released Apr 2024.

What are the best alternatives to Snowflake Arctic Embed M Long?›

E5 Base v2 (better), Gte Qwen2 7B Instruct (open weights).