multilingual-e5-large

multilingual-e5-large is an open-weight multilingual text embedding model that encodes queries and documents into 1024-dimensional dense vectors for semantic search, retrieval-augmented generation, cross-lingual similarity, and clustering across approximately 100 languages. It uses a 24-layer XLM-RoBERTa-large backbone and is part of the intfloat E5 embedding family. Source: supabase.