KURE v1 measured on 13 of 194 Embeddings benchmarks13
BenchmarkByScoreRank
Reranking2
Other11
No benchmark matches
price not verifiedprice
41.1MTEB Thai v1
Dec 2024released
1,024Dimensions
8K tokensMax input
en, koLanguages
Alternatives?
ModelScoreCostPriceWhat changes
SBetterPwC Embedding ExprOpen+5%——Max input: 8K tokens → 1K tokens+1nOpenKURE v2same——sOpenComsat Embed Ko 8B Previewsame——Memory (4-bit): 0.3 GB → 4.5 GBdOpenBGE M3 Kosame——ModelScoreCostPriceWhat changes
Nothing here yet.
ModelScoreCostPriceWhat changes
SPwC Embedding ExprOpen+5%——Max input: 8K tokens → 1K tokens+1Run 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
LicenceMITfree for commercial use
History
3 versionsBenchmarks
in Embeddings & RetrievalFAQ
How much does KURE v1 cost?›
There is no per-unit API price: price not verified.
How good is KURE v1, and what is it best at?›
#64 of 192 in Embeddings.
Is KURE v1 better than KURE v2?›
On MTEB Korean v2 · Reranking · Korean, KURE v1 scores 68.8 vs 69.7 — about the same.
When was KURE v1 released? Is there a newer version?›
Released 2024-12-18. It is the current version in its line.
Is KURE v1 open source?›
Yes, the weights are open (MIT licence).
Can I run KURE v1 locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
What is KURE v1?›
KURE v1 is an open-weights embeddings & retrieval model from nlpai-lab, released Dec 2024.
What are the best alternatives to KURE v1?›
PwC Embedding Expr (better).