I often run visual descriptions through the image analysis feature and then build narratives using gpt‑4 and deepseek. The visual and multimodal support helps me understand complex visuals fast and then write clearly about them. It’s been a great tool for reports.