A mid-sized multimodal model from Google's Gemma family that handles both text and image inputs. It processes visual and textual information together, making it capable of tasks that require understanding images alongside written context. As an open-weight release under Apache 2.0, it can be run and modified locally, though the 'it qat q4 0 unquantized' naming suggests this is an instruction-tuned variant with quantization-aware training details worth noting when deploying.