Alibaba Cloud has introduced a new addition to its Qwen series of artificial intelligence models—the Qwen2.5-Omni-7B, a multimodal large language model engineered to process a wide range of inputs, including text, images, audio, and video. The model can generate real-time text and speech responses, according to a company announcement on Thursday.

As competition in China’s AI space intensifies—especially in the wake of the so-called “DeepSeek moment”—Alibaba’s latest release reinforces its commitment to building accessible, high-performance AI tools.

Designed for deployment on edge devices like smartphones, Qwen2.5-Omni-7B delivers high efficiency without sacrificing performance. “This unique combination makes it an ideal foundation for building agile, cost-effective AI agents—particularly in voice-driven applications,” said Alibaba Cloud.

One potential use case highlighted by the company is assisting visually impaired users with real-time audio descriptions of their surroundings.

The model is now open-sourced and available on platforms like Hugging Face and GitHub, aligning with a growing trend in China to make AI tools more accessible. The move follows DeepSeek’s recent decision to open-source its breakthrough R1 model, which helped spark a new wave of innovation across the country.

Open-source software allows developers to access, modify, and distribute code freely—an approach Alibaba has embraced, having released over 200 generative AI models in recent years.

As China’s AI boom accelerates, major tech firms like Alibaba and Baidu are racing to roll out increasingly sophisticated and cost-effective models. Last week, Baidu unveiled both a new multimodal foundation model and its first model focused on reasoning.

Alibaba, for its part, launched the Qwen 2.5 model in late January and followed it up with an updated version of its AI assistant tool, Quark, earlier this month.

The company has also made bold investments to support its AI ambitions, announcing a $53 billion commitment to cloud computing and AI infrastructure over the next three years—a figure that surpasses its total AI spending over the past decade.

According to Morningstar’s Asia senior equity analyst Kai Wang, Alibaba’s dual focus on building both data centers and proprietary language models positions the company to thrive in China’s post-DeepSeek AI landscape.

Last month, Alibaba scored a major win by confirming a partnership with Apple to integrate its AI technologies into iPhones sold in China. More recently, the company also announced an expanded strategic alliance with BMW to power next-generation AI features in the automaker’s upcoming intelligent vehicles.

Source