Starse measured on 9 of 194 Embeddings benchmarks9
BenchmarkByScoreRank
No benchmark matches
self-host onlyprice
52.5MTEB Russian v1.1
Jun 2026released
ruLanguages
Alternatives?
ModelScoreCostPriceWhat changes
ModelScoreCostPriceWhat changes
Nothing here yet.
ModelScoreCostPriceWhat changes
ModelScoreCostPriceWhat changes
aGiga Embeddings Instruct 10B A1.8B 0826unstatedOpen×2.1——Memory (4-bit): 0 GB → 6.3 GBaGiga Embeddings InstructunstatedOpen×2.1——Memory (4-bit): 0 GB → 1.9 GBaGiga Embeddings Instruct 3B 0826unstatedOpen×2.1——Memory (4-bit): 0 GB → 1.9 GBRun 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
LicenceApache-2.0free for commercial use
Benchmarks
in Embeddings & RetrievalFAQ
How much does Starse cost?›
There is no per-unit API price: self-host only.
Is Starse free to use?›
There is no free API. The weights are open, so you can run it yourself.
How good is Starse, and what is it best at?›
#110 of 194 in Embeddings.
Is Starse better than Paraphrase Multilingual Mpnet Base v2?›
On MTEB Russian v1.1 · Retrieval · Russian, Starse scores 39.1 vs 40.7 — lower.
When was Starse released? Is there a newer version?›
Released 2026-06-25. It is the current version in its line.
Is Starse open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run Starse locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
What is Starse?›
Starse is an open-weights embeddings & retrieval model from BorisTM, released Jun 2026.
What are the best alternatives to Starse?›
EmbeddingGemma 300M (better), Giga Embeddings Instruct 10B A1.8B 0826 (open weights).