A multilingual workhorse built for embedding tasks across dozens of languages. It maps text into dense vector representations, making it useful for semantic search, clustering, and retrieval pipelines. Its strength lies in cross-lingual consistency — similar meanings in different languages tend to land close together in its vector space.