A compact specialist that converts medical text into dense vector representations. With only 300M parameters and a 1024-token context window, it operates efficiently but may struggle with lengthy clinical documents. Its medical domain focus means it likely captures terminology and relationships that general-purpose embedding models might miss.