Alibaba: HappyOyster and its multimodal video generation model HappyHorse are currently undergoing grayscale testing According to TechFlow, on May 13th, Alibaba (09988.HK) reported that we continue to expand the boundaries of AI capabilities through deep innovation. Recently, we have made significant progress in model intelligence by launching a series of new large language models and multimodal models. We released Qwen3.6-Plus in March, achieving significant all-round performance improvement, especially in programming and intelligent agent programming, becoming an industry benchmark in front-end web development and complex warehouse level tasks. Qwen3.6-Plus has stronger multimodal perception and inference capabilities, with a native context window length of up to 1 million tokens, further enhancing stability and credibility. In addition to the Qianwen model family, we also continue to enrich the layout of dedicated models, including the world model HappyOyster that supports real-time creation and interaction, and the multimodal model HappyHorse for video generation. These two models are currently undergoing grayscale testing. (Golden Ten)