Sarashina Embedding v1 1B measured on 7 of 261 Embeddings benchmarks7
BenchmarkByScoreRank
No benchmark matches
self-host onlyprice
81.6JMTEB v2
Nov 2024released
1,792Dimensions
8K tokensMax input
en, jaLanguages
Alternatives?
ModelScoreCostPriceWhat changes
BetterEmbeddingGemma 300MunstatedOpen+4%—$0.002Max input: 8K tokens → 2K tokens+1NOpenJa Embed Sts+5%——Max input: 8K tokens → 1K tokensiBetterE5 Mistral 7B InstructunstatedOpen+2%—$0.022Languages: loses ja+1cBetterF2LLM v2 8BunstatedOpen+5%——Memory (4-bit): 0.7 GB → 4.5 GBcBetterF2LLM v2 14BunstatedOpen+4%——Memory (4-bit): 0.7 GB → 8.4 GBModelScoreCostPriceWhat changes
Nothing here yet.
ModelScoreCostPriceWhat changes
EmbeddingGemma 300MunstatedOpen+4%—$0.002Max input: 8K tokens → 2K tokens+1iE5 Mistral 7B InstructunstatedOpen+2%—$0.022Languages: loses ja+1NJa Embed StsOpen+5%——Max input: 8K tokens → 1K tokenscF2LLM v2 8BunstatedOpen+5%——Memory (4-bit): 0.7 GB → 4.5 GBcF2LLM v2 14BunstatedOpen+4%——Memory (4-bit): 0.7 GB → 8.4 GBModelScoreCostPriceWhat changes
NJa Embed StsOpen+5%——Max input: 8K tokens → 1K tokenscF2LLM v2 8BunstatedOpen+5%——Memory (4-bit): 0.7 GB → 4.5 GB
Harrier OSS v1 0.6BunstatedOpen+4%——cF2LLM v2 4BunstatedOpen+4%——Memory (4-bit): 0.7 GB → 2.4 GBEmbeddingGemma 300MunstatedOpen+4%—$0.002Max input: 8K tokens → 2K tokens+1ModelScoreCostPriceWhat changes
sSarashina Embedding v2 1BOpen+3%——Languages: loses enAll sbintuitions models →Run it yourself
estimateFits 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
Licencehttps://huggingface.co/sbintuitions/sarashina-embedding-v1-1b/blob/main/LICENSEcommercial use with conditions
History
2 versionsBenchmarks
in Embeddings & RetrievalJMTEB v2 · STS · Japanese81.6#28
FAQ
How much does Sarashina Embedding v1 1B cost?›
There is no per-unit API price: self-host only.
Is Sarashina Embedding v1 1B free to use?›
There is no free API. The weights are open, so you can run it yourself.
How good is Sarashina Embedding v1 1B, and what is it best at?›
#28 of 164 in Embeddings.
Is Sarashina Embedding v1 1B better than SFR Embedding 2 R?›
On JMTEB v2 · STS · Japanese, Sarashina Embedding v1 1B scores 81.6 vs 82 — about the same.
When was Sarashina Embedding v1 1B released? Is there a newer version?›
Released 2024-11-22. It is the current version in its line.
Is Sarashina Embedding v1 1B open source?›
Yes, the weights are open (https://huggingface.co/sbintuitions/sarashina-embedding-v1-1b/blob/main/LICENSE licence).
Can I run Sarashina Embedding v1 1B locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
What is Sarashina Embedding v1 1B?›
Sarashina Embedding v1 1B is an open-weights embeddings & retrieval model from sbintuitions, released Nov 2024.
What are the best alternatives to Sarashina Embedding v1 1B?›
EmbeddingGemma 300M (better), Ja Embed Sts (open weights).