Alibaba Qwen-VL has been upgraded, releasing Qwen-VL-Plus and Qwen-VL-Max versions. Qwen-VL-Max surpasses GPT-4V and Gemini in Chinese question answering and text understanding tasks. Qwen-VL-Plus has significantly upgraded detail recognition and text recognition, supporting high pixel resolution. Qwen-VL-Max has improved visual reasoning and instruction-following capabilities, able to recognize GIFs. The new versions demonstrate powerful recognition and reasoning abilities, marking advancements in the field of multimodal models.