Embedding model
A model that turns a passage of text into a single vector, used for search, clustering and retrieval.
Also: text embedding model, sentence encoder
Embedding models are usually encoders trained so that texts with similar meanings produce vectors pointing in similar directions. Sentence-BERT in 2019 was an early widely used example; OpenAI, Google, Cohere, Voyage, Jina and many open projects offer them now, and multilingual ones match a Hindi question to an English passage. They are much smaller and cheaper than chat models. The MTEB benchmark, started by Hugging Face researchers in 2022, compares them across retrieval, classification and other tasks.