AI Tools.

Search
sentence similarity by sentence-transformers

paraphrase-multilingual-MiniLM-L12-v2

Multilingual sentence embedding model covering 50+ languages, built on a 12-layer distilled MiniLM architecture. Produces 384-dimensional vectors designed for semantic similarity and paraphrase detection across language boundaries. Trained on multilingual paraphrase data to align semantically equivalent sentences even when expressed in different languages.

Summary text generated by an automated pipeline from the model card · Not individually reviewed or run by us · How this page is made

Share: X Reddit HN LinkedIn

From the model card

Fields below are copied from the tags and counters on the HuggingFace repository sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2 at our last fetch. They are set by the uploader, not verified by us; rows with no tag are omitted. How this page is made.

Publisher (HF namespace)
sentence-transformers
Pipeline tag
sentence-similarity
Library
Sentence Transformers, Transformers
Framework tags
PyTorch, TensorFlow
Weight formats
ONNX, safetensors, OpenVINO
License tag
apache-2.0 — read the license file in the repo before relying on it
Language tags
multilingual; Arabic (ar), Bulgarian (bg), Catalan (ca), Czech (cs), Danish (da), German (de), Greek (el), English (en), Spanish (es), Estonian (et), Persian (fa), Finnish (fi), French (fr), Galician (gl), Gujarati (gu), Hebrew (he), Hindi (hi), Croatian (hr), Hungarian (hu), Armenian (hy), Indonesian (id), Italian (it), Japanese (ja), Georgian (ka), Korean (ko), Kurdish (ku), Lithuanian (lt), Latvian (lv), Macedonian (mk), Mongolian (mn), Marathi (mr), Malay (ms), Burmese (my), Norwegian Bokmål (nb), Dutch (nl), Polish (pl), Portuguese (pt), Romanian (ro), Russian (ru), Slovak (sk), Slovenian (sl), Albanian (sq), Serbian (sr), Swedish (sv), Thai (th), Turkish (tr), Ukrainian (uk), Urdu (ur), Vietnamese (vi)
Papers cited
arXiv:1908.10084
Downloads (HF counter at last fetch)
45,263,676
Likes (HF counter at last fetch)
1,364
Model card
https://huggingface.co/sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2

Use cases

  • Cross-lingual semantic search (query in one language, docs in another)
  • Multilingual duplicate detection in customer support ticket systems
  • Language-agnostic clustering of community forum posts
  • Building FAQ retrieval for international product lines
  • Paraphrase mining across parallel multilingual corpora

Pros

  • 50+ language coverage in a single model avoids managing per-language checkpoints
  • 384-dim outputs keep vector store costs low relative to 768-dim alternatives
  • Cross-lingual transfer enables single-language labeled data to generalize
  • ONNX and OpenVINO export for production inference; Apache 2.0 license

Cons

  • Smaller distilled architecture limits accuracy vs. per-language specialized models
  • Accuracy gaps between high-resource (en, de, fr) and low-resource languages are significant
  • Shared multilingual tokenizer increases token sequence length for non-Latin scripts
  • 384 dimensions may underfit nuanced semantic distinctions in specialized domains
  • No instruction tuning — prompt phrasing affects embedding quality noticeably

Tags

sentence-transformerspytorchtfonnxsafetensorsopenvinobertfeature-extractionsentence-similaritytransformersmultilingualarbgcacsdadeelenes