DeepSeek-VL

A series of vision-language models released by DeepSeek, supporting multimodal understanding of images and text. It represents DeepSeek's early product line exploring multimodal capabilities.