analiza aiqbench
Alibaba's claims come from a conference speech, so the download and derivative-model figures are company-reported and not independently audited. Still, they point to a real trend: open-weight models from Chinese developers are gaining traction in the global developer ecosystem. Readers should watch whether download counts translate into production deployments and independent benchmark results, since metrics like downloads and derivatives measure reach more than capability.
Alibaba's Joseph Tsai: Open Source Drives AI Adoption, Qwen Open-Source Models Surpass 4 Billion Cumulative Downloads Worldwide
On the morning of November 3, the 2026 World Internet Conference Wuzhen Summit opened. Joseph Tsai, Alibaba Group co-founder and chairman of the board, spoke at the opening ceremony, saying that AI can only be inclusive and widely accessible if it is open source and open. Chinese open-source models offer strong overall performance and high inference efficiency, with key metrics such as download volume and the number of derivative models remaining in the lead, making them the current global leaders in open-source intelligence. Alibaba's Qwen has open-sourced more than 470 models, with cumulative global downloads exceeding 4 billion and over 330,000 derivative models, ranking first globally.
Citește articolul complet pe 36Kr