A compact, specialized vision-language model fine-tuned specifically for dense, detailed image captioning. It punches above its weight for describing visual content with rich, structured language, but its narrow fine-tuning means it performs best when the task closely resembles detailed image description rather than general-purpose conversation or reasoning.