# Cross-lingual Retrieval

A sentence transformer model supporting multiple languages, suitable for text ranking tasks

A 107M-parameter multilingual embedding model developed by IBM, supporting 12 languages with excellent performance on the MTEB benchmark

This is an intermediate version of a multilingual Mixture of Experts (MoE) text embedding model, obtained through multi-stage contrastive training

Multilingual E5 large instruction model supporting text embedding and classification tasks in over 100 languages

A bge-m3 model optimized for English and Russian, featuring a streamlined vocabulary that retains only English and Russian words, reducing the vocabulary size to 21% of the original while maintaining the quality of embeddings for both languages. The total model parameters are 63.3% of the original version.