Liang Wang, Nan Yang, Xiaolong Huang, Linjun Yang, Rangan Majumder, Furu Wei, arXiv 2024 This model has 12 layers and the embedding size is 384. Below is an example to encode queries and passages from the MS-MARCO passage ranking dataset. This model is initialized from microsoft/Multilingual-MiniLM-L12-H384 and continually trained on a mixture of multilingual datasets. It supports 100 languages from xlm-roberta, but low-resource languages may see performance degradation. For all labeled datasets, we only use its training set for fine-tuning. For other training details, please refer to our paper at https://arxiv.org/pdf/2402.05672. Check out unilm/e5 to reproduce evaluation results on the…
Open-weight model · Sentence similarity
paraphrase-multilingual-MiniLM-L12-v2
by Sentence Transformers sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
This is a sentence-transformers model: It maps sentences & paragraphs to a 384 dimensional dense vector space and can be used for tasks like clustering or semantic search.
Parameters118M
Context512
Weights4.6 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads45.7M
Runs On
What it takes to serve paraphrase-multilingual-MiniLM-L12-v2 (118M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.
| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
|---|---|---|---|---|---|
| 16-bit | 0.2 GB | 0.3 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 0.1 GB | 0.1 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 0.1 GB | 0.1 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026.
SAVRN's Notes on paraphrase-multilingual-MiniLM-L12-v2
Every sentence or paragraph you feed this comes back as 384 numbers, a dense vector you can cluster or search against. The vocabulary runs to 250,037 entries and the name says multilingual, so one index can hold more than one language. Memory is not a decision here: 0.3 GB at 16-bit on the cheapest card we track, a 192 GB MI300X at $1.85 an hour, so it shares a card rather than owning one.
Check the 512-token ceiling first: anything longer gets chunked before it is embedded. The last update landed on 2026-01-28, so pin the revision your pipeline was built against. Without the sentence-transformers library you run the transformer and apply the pooling step yourself. Apache 2.0 permits commercial use, modification and redistribution; keep the notices, note significant changes, and the patent grant from contributors comes with it.
Model Card
By Sentence Transformers, published under apache-2.0, revision e8f8c211226b.
This is a sentence-transformers model: It maps sentences & paragraphs to a 384 dimensional dense vector space and can be used for tasks like clustering or semantic search.
Usage (Sentence-Transformers)
Using this model becomes easy when you have sentence-transformers installed:
pip install -U sentence-transformers
Then you can use the model like this:
from sentence_transformers import SentenceTransformer
sentences = ["This is an example sentence", "Each sentence is converted"]
model = SentenceTransformer('sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2')
embeddings = model.encode(sentences)
print(embeddings)
Usage (HuggingFace Transformers)
Without sentence-transformers, you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.
Configuration
- Architecture
- BertModel
- Context length (tokens)
- 512
- Layers
- 12
- Hidden size
- 384
- Feed-forward size
- 1,536
- Attention heads
- 12
- Vocabulary size
- 250,037
- Model type
- bert
Identity and Version
- Repository
- sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
- Publisher
- Sentence Transformers
- Task
- Sentence similarity
- Modality
- Text
- Library
- sentence-transformers
- Parameters
- 118M parameters
- Languages
- ar, bg, ca, cs, da, de, el, en
- Revision
- e8f8c211226b894fcb81acc59f3b34ba3efd5f42
- First published
- 2022-03-02
- Last updated
- 2026-01-28
Files and Weights
28 files, 4.6 GB in total. The weights are 14 files totalling 4.6 GB in bin, h5, onnx, safetensors.
Weights14 files · 4.6 GB
Configuration7 files · 14.8 MB
Tokenizer2 files · 9.1 MB
Documentation1 file · 3.9 KB
Other3 files · 6.2 MB
Repository1 file · 1.0 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model.safetensors | Weights | 470.6 MB | eaa086f0ffee |
| onnx/model.onnx | Weights | 470.3 MB | 10f7a0884202 |
| onnx/model_O1.onnx | Weights | 470.2 MB | 9ae4b831e992 |
| onnx/model_O2.onnx | Weights | 470.1 MB | 338ef03c2838 |
| onnx/model_O3.onnx | Weights | 470.1 MB | 2426785672da |
| onnx/model_O4.onnx | Weights | 235.2 MB | 307bba13f9f5 |
| onnx/model_qint8_arm64.onnx | Weights | 118.4 MB | 783fea82d71a |
| onnx/model_qint8_avx512.onnx | Weights | 118.4 MB | 783fea82d71a |
| onnx/model_qint8_avx512_vnni.onnx | Weights | 118.4 MB | 783fea82d71a |
| onnx/model_quint8_avx2.onnx | Weights | 118.5 MB | 98a01d88b7de |
| openvino/openvino_model.bin | Weights | 470.0 MB | 04e8cc4ceedb |
| openvino/openvino_model_qint8_quantized.bin | Weights | 119.0 MB | 24acd56a5f5a |
| pytorch_model.bin | Weights | 470.7 MB | 16cc9e54df6e |
| tf_model.h5 | Weights | 470.9 MB | 22150b6ba00e |
| 1_Pooling/config.json | Configuration | 190 B | — |
| config.json | Configuration | 645 B | — |
| config_sentence_transformers.json | Configuration | 122 B | — |
| modules.json | Configuration | 229 B | — |
| sentence_bert_config.json | Configuration | 53 B | — |
| special_tokens_map.json | Configuration | 239 B | — |
| unigram.json | Configuration | 14.8 MB | 71b44701d7ef |
| README.md | Documentation | 3.9 KB | — |
| openvino/openvino_model.xml | Other | 398.5 KB | — |
| openvino/openvino_model_qint8_quantized.xml | Other | 708.8 KB | — |
| sentencepiece.bpe.model | Other | 5.1 MB | cfc8146abe2a |
| .gitattributes | Repository | 1.0 KB | — |
| tokenizer.json | Tokenizer | 9.1 MB | 2c3387be7655 |
| tokenizer_config.json | Tokenizer | 526 B | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 4.6 GB
Download from Sentence Transformers
Released by Sentence Transformers through its official repository on Hugging Face. Read the license.
Built From
- Described by arXiv:1908.10084
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 4.6 GB |
| 16-bit | 0.2 GB |
| 8-bit | 0.1 GB |
| 4-bit | 0.1 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Compare paraphrase-multilingual-MiniLM-L12-v2
Questions About paraphrase-multilingual-MiniLM-L12-v2
How much GPU memory does paraphrase-multilingual-MiniLM-L12-v2 need?
About 0.3 GB at 16-bit and 0.1 GB at 4-bit: the weights (118M parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run paraphrase-multilingual-MiniLM-L12-v2 on?
At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
Can I use paraphrase-multilingual-MiniLM-L12-v2 commercially?
Yes. paraphrase-multilingual-MiniLM-L12-v2 is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.
What is paraphrase-multilingual-MiniLM-L12-v2's context length?
512 tokens, from the maximum position embeddings in its published configuration.
Similar Models
This is a LinkTransformer model. At its core this model this is a sentence transformer model sentence-transformers model- it just wraps around the class. It is designed for quick and easy record linkage (entity-matching) through the LinkTransformer package. The tasks include clustering, deduplication, linking, aggregation and more. Notwithstanding that, it can be used for any sentence similarity task within the sentence-transformers framework as well. It maps sentences & paragraphs to a 384 dimensional dense vector space and can be used for tasks like clustering or semantic search. Take a look at the documentation of sentence-transformers if you want to use this model for more than what we…
This is a sentence-transformers model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search. Using this model becomes easy when you have sentence-transformers installed: Then you can use the model like this: Without sentence-transformers, you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings. KorSTS, KorNLI 학습 데이터셋으로 멀티 태스크 학습을 진행한 후 KorSTS 평가 데이터셋으로 평가한 결과입니다. The model was trained with the parameters: sentencetransformers.datasets.NoDuplicatesDataLoader.NoDuplicatesDataLoader of…
This is a LinkTransformer model. At its core this model this is a sentence transformer model sentence-transformers model- it just wraps around the class. It is designed for quick and easy record linkage (entity-matching) through the LinkTransformer package. The tasks include clustering, deduplication, linking, aggregation and more. Notwithstanding that, it can be used for any sentence similarity task within the sentence-transformers framework as well. It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search. Take a look at the documentation of sentence-transformers if you want to use this model for more than what we…
This is a LinkTransformer model. At its core this model this is a sentence transformer model sentence-transformers model- it just wraps around the class. It is designed for quick and easy record linkage (entity-matching) through the LinkTransformer package. The tasks include clustering, deduplication, linking, aggregation and more. Notwithstanding that, it can be used for any sentence similarity task within the sentence-transformers framework as well. It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search. Take a look at the documentation of sentence-transformers if you want to use this model for more than what we…
This is a LinkTransformer model. At its core this model this is a sentence transformer model sentence-transformers model- it just wraps around the class. It is designed for quick and easy record linkage (entity-matching) through the LinkTransformer package. The tasks include clustering, deduplication, linking, aggregation and more. Notwithstanding that, it can be used for any sentence similarity task within the sentence-transformers framework as well. It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search. Take a look at the documentation of sentence-transformers if you want to use this model for more than what we…