MTEB logo MTEB
Japanese icon

Japanese

Cite this benchmark

@inproceedings{li-etal-2026-jmteb,
  address = {Palma de Mallorca, Spain},
  author = {Li, Shengzhe and Ohagi, Masaya and Ri, Ryokan and Fukuchi, Akihiko and Shibata, Tomohide and Kawahara, Daisuke},
  booktitle = {Proceedings of the Fifteenth Language Resources and Evaluation Conference},
  doi = {10.63317/5ouzpv2f2f6k},
  month = may,
  pages = {7423--7434},
  publisher = {ELRA Language Resource Association},
  title = {{JMTEB} and {JMTEB}-lite: {J}apanese Massive Text Embedding Benchmark and Its Lightweight Version},
  url = {https://aclanthology.org/2026.lrec-1.588/},
  year = {2026},
}

Japanese text embedding quality across clustering, classification, semantic similarity, retrieval, and reranking. v2 extends the benchmark to 28 datasets for more comprehensive evaluation compared with MTEB(jpn, v1).

Reference paper →
Languages 1
Tasks 28
Task Types 5
Models 0