Model reference · open weights

starse

starse is an open-weight embedding model from BorisTM, listed in the AxForge catalogue. AxForge can bring it up on EU-owned hardware for you on request — with the licence handled where one is required.

Embeddings BorisTM 1 variants 556 downloads/mo
Request this model on EU hardware All served models Not on the shared API today — deployed on request.

About

What starse is

StaRSE StaRSE stands for Static Russian Sentence Embeddings. It is a compact Russian sentence embedding model implemented as a Sentence-Transformers StaticEmbedding endpoint. The model is intended for CPU-friendly semantic similarity, clustering, classification features, and retrieval-style first-stage representations when a full Transformer encoder is too expensive to run at high throughput. [!NOTE] StaRSE has 61.51M logical embedding parameters (120,138 × 512), stored as binary sign bits plus one FP32 L2 norm vector. Performance Evaluation is reported on MTEB(rus, v1.1) across 23 tasks. The main score is meantaskmainscore = 51.16. Usage Install Sentence Transformers: Load the model with trustremotecode=True. Citation

Summarised from the published model card. Read the full card on the HuggingFace links below.

Specifications

What it is

MakerBorisTM
TypeEmbedding models
Parameters (lead)8M
Variants1
Runs withsentence-transformers
Released2026-05-23
Popularity556 downloads / month
Likes4
LicenceOpen weights

How it works

How embedding models work

Your textsentence / documentEncodermaps meaningVectorlist of numbersAn embedding model turns text into a vector, so similar meanings sit close together — the basis of search and RAG.

Variants

Sizes & precisions

Open weights ship in several sizes and precisions. One page, all the variants — pick the one that fits your GPU. VRAM figures are estimates from model size.

VariantParamsPrecisionVRAMFits 16 GBWeights
starse8MBF16~0 GBWeights ↗

Using it via the API

Call it like any OpenAI endpoint

Once AxForge deploys starse for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (starse below is illustrative; you get the exact model name on deployment.)

$ curl -sS https://api.axforge.ai/v1/embeddings \
  -H "Authorization: Bearer $AXFORGE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"starse","input":"text to embed"}'

Details

Languages, data & research

Languages

ru

Trained / evaluated on

unicamp-dl/mmarco deepvk/ru-HNP deepvk/cultura_ru_edu cointegrated/ru-paraphrase-NMT-Leipzig

Tags

sentence-transformers safetensors sentence-similarity feature-extraction static-embeddings binary russian ru dataset:unicamp-dl/mmarco dataset:deepvk/ru-HNP dataset:deepvk/cultura_ru_edu dataset:cointegrated/ru-paraphrase-NMT-Leipzig endpoints_compatible 8-bit

Licence

Open weights

Open weights under apache-2.0 — commercial use is permitted. Deploy it on AxForge EU hardware on request. Read the licence ↗

Sources

Weights & code

Want starse on EU-owned hardware?

Request this model on EU hardware See what’s served now

Explore

More embedding models

© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms