XrayGLM is a vision-language model built on VisualGLM-6B, fine-tuned to interpret medical chest radiographs and produce diagnostic summaries in Chinese. It adapts general-domain multimodal LLMs like MiniGPT-4 and LLaVA to the medical domain, addressing the lack of Chinese-language medical VLM research. The project targets healthcare AI applications requiring visual understanding combined with natural language generation.