A compact, no-frills text encoder that converts sentences and passages into dense numerical vectors. It handles general English text reliably, making it a practical workhorse for semantic search and similarity tasks. Its modest size keeps inference fast and resource-light, though it trades some nuance for that efficiency.