Up to date
sentence-transformers Apache 2.0 out 5 years ago
278M parameters, text in, embeddings out, 514 tokens of context. 1 provider serves it.
Verified 2026-10-25 Rank #19 in Embeddings · score 99.5 method v1 Sources 3 Downloads 30d 8.7M · #12 of 619 tracked
Picked by rule from 5 tracked pieces: the lab's own post, the most shared, a field report, the latest.
The paraphrase-multilingual-mpnet-base-v2 model converts sentences into 768-dimensional vectors for semantic tasks across 50+ languages.
The 1 pick above, plus these 4. Top is the radar's score: discussion, source authority and the lab's own posts, fading with age.
Live hosted endpoints, cheapest first, USD per million tokens. Ties break on output price, then provider name.
| # | Cheapest live row | In /1M | Out /1M | Context | Quant | Seen |
|---|---|---|---|---|---|---|
| 1 | HF Inference | — | — | — | — | Oct 25 |
Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.
Served by 1 API provider; 17 quantized builds for local use; a big base for builders (183 fine-tunes, adapters and merges).
Top Spaces running it: cvachet/pdf-chatbot ♥ 382 · embedl/hfviewer ♥ 181 · IlyaGusev/saiga_13b_llamacpp_retrieval_qa ♥ 47 · myscale/ChatData ♥ 23
Weights only, an estimate: leave room for the KV cache. Method
About 0.2 GB at INT4, 0.3 GB at FP8, 0.7 GB at BF16: weights only, so leave room for the KV cache.
Estimate params × bytes per param × 1.2 · how we estimate
Third-party measurements only; author-reported numbers stay in Reference, marked as such.
58.27. Each tick is a model on this site with a measured MTEB row; author-reported numbers stay in Reference. Mean(task) x100 over 41 tasks. 41 tasks, split=test
Source · leaderboard · 2026-10-24
Weights, every dated article, and repo commits on one rail.
Teal is the weights (2022-03-02), amber triangles are 1 dated article (each links to it), grey ticks are repo commits.
200 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Everything else, folded: the full tables, every source.
| Benchmark | Value | Setting / notes | Reported by | Date |
|---|---|---|---|---|
| measured MTEB · Embed text for search | 37.14 nDCG@10 x100 (mean of 10 tasks) | 10 tasks, split=test | leaderboard | 2026-10-24 |
| measured MTEB · Embed text for search | 58.27 mean(task) x100 over 41 tasks | 41 tasks, split=test | leaderboard | 2026-10-24 |
| measured MTEB Reranking · Rerank search results | 45.16 MAP x100 (mean of 2 tasks) | 2 tasks, split=test | leaderboard | 2026-10-24 |
Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.
Listed in Embeddings
| Pipeline tag | sentence-similarity |
|---|---|
| Library | sentence-transformers |
| Architectures | XLMRobertaModel |
| Model type | xlm-roberta |
| Params by dtype | F32: 278,043,648 · I64: 514 |
| Size band | S |
| MoE | no |
| Context length | 514 (config.json max_position_embeddings) |
| Languages | multilingual, ar, bg, ca, cs, da, de, el, en, es, et, fa, fi, fr, gl, gu, he, hi, hr, hu, hy, id, it, ja, ka, ko, ku, lt, lv, mk, mn, mr, ms, my, nb, nl, pl, pt, ro, ru, sk, sl, sq, sr, sv, th, tr, uk, ur, vi |
| Papers | arXiv:1908.10084 |
| License | Apache 2.0 (apache-2.0) |
| Hub snapshot | 2026-10-25 |
vLLM: supported · XLMRobertaModel · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · XLMRobertaModel · 773cb938fe9484a8f7e7e46fb1b1fe96412c59ed
vLLM registry (7050ef881ff0835f6890a47db1859a24c65481d6) · SGLang registry (773cb938fe9484a8f7e7e46fb1b1fe96412c59ed) · Matched by architecture class name against each project's registry on main; 'supported' means the architecture is registered, not that this checkpoint was tested.
Weights-only memory for paraphrase-multilingual-mpnet-base-v2: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 0.7 GB · Edge (<4 GB) |
|---|---|
| FP8 (1 byte/param) | 0.3 GB · Edge (<4 GB) |
| INT4 (0.5 bytes/param) | 0.2 GB · Edge (<4 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.
finetune: 183quantized: 17
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| TechWolf/JobBERT-v3 | finetune | — | 278M | S | mit | TechWolf | 28 | 652.1k |
| Xenova/paraphrase-multilingual-mpnet-base-v2 | quantized | — | — | — | — | Xenova | 4 | 563k |
| Gameselo/STS-multilingual-mpnet-base-v2 | finetune | — | 278M | S | — | Gameselo | 3 | 36.1k |
| Omartificial-Intelligence-Space/Arabic-all-nli-triplet-Matryoshka | finetune | — | 278M | S | apache-2.0 | Omartificial-Intelligence-Space | 6 | 13.7k |
| minhquan6203/paraphrase-vietnamese-law | finetune | — | 278M | S | — | minhquan6203 | 1 | 5.4k |
| mykor/paraphrase-multilingual-mpnet-base-v2.gguf | quantized | gguf | — | — | apache-2.0 | mykor | 0 | 3.6k |
| GrishaKushnir/jobbert-onnx | quantized | — | — | — | mit | GrishaKushnir | 0 | 3.4k |
| cruciverb-it/crossword-space-mpnet-base-ade | finetune | — | 557M | S | cc-by-4.0 | cruciverb-it | 3 | 2.8k |
| ODeNy/ChecketV2 | finetune | — | 278M | S | apache-2.0 | ODeNy | 0 | 2k |
| hassan7272/urdu-finance-embeddings | finetune | — | 278M | S | apache-2.0 | hassan7272 | 0 | 1.4k |
| MesTruck/STS-multilingual-mpnet-base-v2-GGUF | quantized | GGUF | — | — | — | MesTruck | 0 | 1.2k |
| yahyaabd/allstats-semantic-search-model-v1-3 | finetune | — | 278M | S | — | yahyaabd | 11 | 421 |
| gentilrenard/paraphrase-multilingual-mpnet-base-v2_setfit-lemonde-french | finetune | — | 278M | S | apache-2.0 | gentilrenard | 0 | 911 |
| impresso-project/histlux-paraphrase-multilingual-mpnet-base-v2 | finetune | — | 278M | S | agpl-3.0 | impresso-project | 0 | 811 |
| sizrox/paraphrase-multilingual-mpnet-base-v2-Q8_0-GGUF | quantized | Q8_0 | — | — | apache-2.0 | sizrox | 1 | 685 |
| gmunkhtur/finetuned_tdb_paraphrase-multilingual_mpnet_try6 | finetune | — | 278M | S | — | gmunkhtur | 1 | 577 |
| soof/miswag-category-mapper | finetune | — | 278M | S | — | soof | 0 | 554 |
| l0rdkr0n0s/albanian-legal-sentence-transformer | quantized | — | 278M | S | apache-2.0 | l0rdkr0n0s | 1 | 496 |
| Kowshik24/bangla-sentence-transformer-ft-matryoshka-paraphrase-multilingual-mpnet-base-v2 | finetune | — | 278M | S | apache-2.0 | Kowshik24 | 0 | 540 |
| SANIAAFTAR/bge-base-en-v1.5-finetuned_ragds_v2 | finetune | — | 278M | S | — | SANIAAFTAR | 1 | 474 |
| saraleivam/GURU-model2 | finetune | — | 278M | S | — | saraleivam | 0 | 521 |
| hakansabunis/trakad-embed-v2 | finetune | — | 278M | S | apache-2.0 | hakansabunis | 0 | 509 |
| SANIAAFTAR/bge-base-en-v1.5-finetuned_ragds_v1 | finetune | — | 278M | S | — | SANIAAFTAR | 0 | 505 |
| mranonymaz/bol-topic-classifier | finetune | — | 278M | S | — | mranonymaz | 0 | 487 |
| Yaaasmineee/sbert-cv-job-matching | finetune | — | 278M | S | — | Yaaasmineee | 1 | 437 |
| abdulrehman1998/bge-base-en-v1.5-finetuned_ragds_v1 | finetune | — | 278M | S | — | abdulrehman1998 | 0 | 463 |
| olgapoletkina/fine-tuned-nuf-room-classifier | finetune | — | 278M | S | — | olgapoletkina | 0 | 425 |
| aloizidis/make-multilingual-en-lb-checkpoint-5000 | finetune | — | 278M | S | — | aloizidis | 0 | 423 |
| IvanDVonga/LegalEmbed | finetune | — | 278M | S | cc-by-4.0 | IvanDVonga | 0 | 420 |
| Sayyam-1/crislens-cross-mpnet | finetune | — | 278M | S | — | Sayyam-1 | 0 | 417 |
| IoannisKat1/paraphrase-multilingual-mpnet-base-v2-matryoshka1 | finetune | — | 278M | S | apache-2.0 | IoannisKat1 | 0 | 394 |
| ppsingh/mpnet-multi-agri-classifier | finetune | — | 278M | S | apache-2.0 | ppsingh | 0 | 388 |
| RomainDarous/large_directOneEpoch_meanPooling_mistranslationModel | finetune | — | 278M | S | — | RomainDarous | 0 | 385 |
| Sayyam-1/crislens-mono-mpnet | finetune | — | 278M | S | — | Sayyam-1 | 0 | 383 |
| inesctec/CitiLink-mpnet-Entity-Linker-pt | finetune | — | 278M | S | other | inesctec | 0 | 379 |
| pyrac/rse_engagement_des_collaborateurs | finetune | — | 278M | S | apache-2.0 | pyrac | 0 | 378 |
| yahyaabd/allstats-search-base-v1-64-1 | finetune | — | 278M | S | — | yahyaabd | 0 | 372 |
| gmunkhtur/finetuned_tdb_paraphrase-multilingual_mpnet_00 | finetune | — | 278M | S | — | gmunkhtur | 0 | 370 |
| pocohos/paraphrase-multilingual-mpnet-base-v2-Q6_K-GGUF | quantized | Q6_K | — | — | apache-2.0 | pocohos | 0 | 362 |
| saraleivam/GURU-model | finetune | — | 278M | S | — | saraleivam | 0 | 360 |
| Date | Commit | SHA |
|---|---|---|
| 2025-08-19 | Add `text-embeddings-inference` tag & snippet (#14) | 4328cf2 |
| 2025-03-06 | Remove deprecated (SEB) evaluation results section | 84fccfe |
| 2024-11-05 | Add exported openvino model 'openvino_model_qint8_quantized.xml' | 75c5775 |
| 2024-10-11 | Add new SentenceTransformer model with an openvino backend | adbb5d9 |
| 2024-10-11 | Add new SentenceTransformer model with an openvino backend | 030d413 |
| 2024-10-11 | Add exported ONNX model 'model_qint8_avx512_vnni.onnx' | 00a11c7 |
| 2024-10-11 | Add exported ONNX model 'model_qint8_avx512.onnx' | f8a4354 |
| 2024-10-11 | Add exported ONNX model 'model_quint8_avx2.onnx' | 97fe4ea |
| 2024-10-11 | Add exported ONNX model 'model_qint8_arm64.onnx' | 8033eee |
| 2024-10-11 | Add exported ONNX model 'model_O4.onnx' | 088a2c0 |
| 2024-10-11 | Add exported ONNX model 'model_O3.onnx' | ac47576 |
| 2024-10-11 | Add exported ONNX model 'model_O2.onnx' | bb931c0 |
| 2024-10-11 | Add exported ONNX model 'model_O1.onnx' | 2f306b1 |
| 2024-10-11 | Add new SentenceTransformer model with an onnx backend | 4840f22 |
| 2024-03-27 | Adding `safetensors` variant of this model (#7) | 79f2382 |
| 2024-02-12 | Update Sentence Transformers metadata (#11) | 33a22e7 |
| 2023-11-02 | replaced max with avg in comment (#8) | 4e754b7 |
| 2023-11-02 | Update README.md (#1) | 67b8354 |
| 2022-06-15 | Add TF weights | ef15aed |
| 2021-08-05 | Update README.md | fa41ae1 |
| 2021-06-23 | Add new SentenceTransformer model. | af122d2 |
| 2021-06-02 | upload | a2a36cb |
| 2021-06-02 | initial commit | 0d1a7a7 |
Hosted endpoints observed for paraphrase-multilingual-mpnet-base-v2, sorted by input price (USD per million tokens). Each row names its observer and date; prices change daily, so check the source before quoting.
| Provider | Model ref | Context | Quant | $/M in | $/M out | tok/s | TTFT | Tools | Status | Observed |
|---|---|---|---|---|---|---|---|---|---|---|
| HF Inference hf | sentence-transformers/paraphrase-multilingual-mpnet-base-v2 | — | — | — | — | — | — | — | live | 2026-10-25 |
Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency. Summaries are generated from a stored copy of the article text; every sentence cites a span of that text, and nothing is shown that failed the citation check.
arXiv links found in the model card and its tags. "Peer-reviewed" is shown only with evidence: a published venue in OpenAlex (measured) or the authors' own journal reference on arXiv (reported).
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:33 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:45 UTC · Summary: 2026-10-07 15:00 UTC · Last verified 2026-10-25
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.
