TechFlow News, May 13: Alibaba Group (09988.HK) reported in its financial results that it continues to expand the boundaries of AI capabilities through deep innovation. Recently, the company has achieved significant progress in model intelligence by launching a series of new large language models and multimodal models. In March, Alibaba released Qwen3.6-Plus, delivering substantial all-around performance improvements—particularly excelling in programming and agent-based programming, establishing itself as an industry benchmark for front-end web development and complex repository-level tasks.
Qwen3.6-Plus also features enhanced multimodal perception and reasoning capabilities, with a native context window length supporting up to 1 million tokens, further improving stability and reliability. Beyond the Qwen model family, Alibaba continues to enrich its portfolio of specialized models, including HappyOyster—a world model enabling real-time creation and interaction—and HappyHorse—a multimodal model for video generation. Both models are currently undergoing gray-scale testing. (Jinshi)
