A compact multimodal model from LiquidAI that handles both text and image inputs. At 2B parameters, it sits at the smaller end of the spectrum, making it lightweight and deployable in resource-constrained environments. Specific capability details beyond its multimodal input support are limited.