A compact vision-language model that handles both text and image inputs, producing text outputs. At 8 billion parameters, it sits in the mid-range tier where capability and resource efficiency intersect. Details about its specific strengths, quirks, and behavioral tendencies are limited, making it difficult to characterize its working style beyond its multimodal input support.