Gemma 4 is Google's open-weight text model that brings capable language understanding to environments where local deployment matters. It carries the Gemma family's reputation for being relatively compact and accessible, making it practical for developers who want to run inference without cloud dependencies. Like its predecessors, it trades some raw capability headroom for deployability and transparency.