A compact, focused embedding specialist built for efficiency over breadth. With only a 512-token context window, it works best on short, targeted text snippets rather than long documents. Its MiniLM lineage means it trades raw capacity for speed and low resource consumption.