March 27 - Alibaba (BABA, Financial) has launched a new artificial intelligence model under its Qwen series as competition in China's AI sector accelerates.

The Qwen2.5-Omni-7B model is a unified multimodal system capable of processing text, images, audio, and video inputs while generating real-time text and speech outputs. The model follows the company's earlier release of the updated Qwen 2.5 version in January.

The company said the model, designed with 7 billion parameters, delivers strong performance despite its compact size. It is open-sourced and available on platforms such as Hugging Face and GitHub.

Alibaba said the model is optimized for deployment on edge devices such as smartphones and laptops. The company expects it to support cost-effective AI agents, including applications for visually impaired users via real-time audio guidance.

The launch comes amid growing competition from Chinese peers such as Baidu (BIDU, Financial) and Tencent (TCEHY, Financial), which are also expanding their AI capabilities following the introduction of DeepSeek's R1 model last year.

Alibaba also reaffirmed its commitment to AI investment, with plans to allocate $53 billion toward AI infrastructure and cloud computing over the next three years.

Baidu earlier this month advanced its AI roadmap by launching upgraded reasoning models, while Alibaba updated its AI-powered assistant tool, Quark.

Alibaba shares have climbed about 55% year-to-date, supported by positive sentiment around its AI expansion and improving financial outlook.