According to DeepLearning.AI, Alibaba has launched the Qwen2.5-VL, a series of open vision-language models with parameter sizes of 3B, 7B, and 72B. These models are capable of processing images, text, and videos, and can interact with computer interfaces. The largest model, Qwen2.5-VL-72B, surpassed competitors like GPT-4o and Gemini, indicating a significant advancement in AI capabilities. This could influence trading strategies in AI-focused tech stocks as Alibaba continues to innovate in the AI space.