At the 2025 World Artificial Intelligence Conference (WAIC), Alibaba introduced three brand-new open-source large models, a move that undoubtedly injected a strong boost into the global AI field. The models released this time cover foundational models, programming models, and reasoning models, all of which have won the highest honors in the global open-source community, marking Alibaba's strong capabilities in open-source technology.
The most notable among them is the "Qwen" model. This model achieves full-scale, multi-modal open-sourcing and has become the most downloaded and the most derived open-source model family globally. It can be imagined that future developers will gain more rich and diverse tools when using these models, thus promoting the rapid development of various AI applications.
In addition to the release of new models, Alibaba also demonstrated its "full-stack AI" capabilities, aiming to improve the efficiency of enterprises and developers by fully integrating cloud computing and artificial intelligence. To achieve this goal, Alibaba plans to invest up to 38 billion yuan over the next three years for the construction of cloud and AI infrastructure. This ambitious plan will further strengthen Alibaba's leading position in the AI field.
Notably, Alibaba Cloud's Hundredfold platform was selected as a "treasure of the museum," with more than 200,000 developers gathered on the platform, who have developed over 700,000 intelligent agents. This data not only reflects Alibaba Cloud's influence in the industry but also demonstrates the appeal of its platform to developers.
With the release of these open-source large models, Alibaba has undoubtedly taken a more significant position in the global AI industry. We look forward to these new technologies bringing more innovation and opportunities to various industries!