Alibaba Group Holding’s new Qwen3-Omni multimodal artificial intelligence system has quickly become the most popular model in the world’s largest open-source AI community, challenging closed systems from OpenAI and Google while underscoring the growing popularity of Chinese open AI systems.
Qwen3-Omni-30B-A3B, a variant from the Qwen3-Omni series, ranked top on Hugging Face’s trending model list as of Monday, followed by image editing model Qwen-Image-Edit-2509. Alibaba owns the Post.
Alibaba’s cloud computing unit unveiled the Qwen3-Omni series on Tuesday, describing it as the first native end-to-end multimodal system that “unifies text, images, audio and video in one model”.
Shares of Alibaba rose 4.14 per cent in Hong Kong on Monday, closing at HK$173.40.
Initially launched in April, Qwen3 has expanded into a family of models possessing text, image, audio and visual capabilities, as well as various multimodal systems tailored for real-world applications.
[LIVE] China Future Tech webinar | A deep dive into China’s open-source AI revolution
[LIVE] China Future Tech webinar | A deep dive into China’s open-source AI revolution
Qwen researcher Lin Junyang attributed the new multimodal system’s enhancements to general improvements across foundational projects related to audio and images. The Qwen team “combined everything … to build our Qwen3-Omni”, he said.