Alibaba has unveiled a set of new open-source artificial intelligence models and text-to-video AI technology, ramping up its initiatives to compete in the rapidly expanding field of generative AI. Over 100 open-source models have been released from Alibaba’s Qwen 2.5 family, which is its latest foundational large language model introduced in May.
While competitors like Baidu opened new tab and OpenAI have mainly worked on closed-source strategies, Alibaba has equally combined both proprietary and open-source models to expand its AI product portfolio.
The newest models vary from 0.5 to 72 billion parameters the factors that define the prowess and efficiency of an AI model – in terms of size, for mathematics, coding skills, and multilingual support of over 29 languages, Alibaba announced.
The models are designed to support a diverse range of AI applications in sectors like automotive, gaming, and scientific research. Alibaba introduced a new text-to-video model as part of its Tongyi Wanxiang image generation series, aligning itself with a rising trend among Chinese tech companies entering this burgeoning market. This development positions Alibaba to compete directly with international firms such as OpenAI, which is also exploring text-to-video technology.