Alibaba announced its new artificial intelligence model, Qwen 3.5, on Monday. The company states that the model is capable of independently performing complex tasks and provides significant improvements in efficiency and cost compared to previous versions. Alibaba asserts that Qwen 3.5 surpasses leading AI models from the United States in various benchmark tests.
The introduction of this model aims to broaden the user base for Alibaba's Qwen chatbot application within China. The Chinese AI market is currently characterized by strong competition, with companies like ByteDance's Doubao and DeepSeek, which achieved an early global recognition, being prominent players.
According to Alibaba, Qwen 3.5 is considerably more cost-effective, with operational expenses reduced by 60%, and it demonstrates an eight-fold increase in efficiency when handling substantial data loads compared to its predecessor. A key feature highlighted is its "visual agentic capabilities," which enable the model to autonomously interact with and control mobile and desktop applications.
In a statement, Alibaba explained that Qwen 3.5 is developed for the "agentic AI era," intended to empower developers and businesses to accelerate their operations and achieve more with the same computational resources, thereby establishing a new standard for capability relative to inference cost.
Meanwhile, ByteDance released Doubao 2.0 on Saturday, an updated version of its chatbot app which currently serves close to 200 million users. This release also emphasizes its suitability for the developing AI agent landscape.
Industry observers suggest that the deployment of Qwen 3.5 could bolster Alibaba's position in China's highly competitive AI sector. Earlier in February, the e-commerce firm experienced a sevenfold surge in active users for its Qwen chatbot following a promotional coupon campaign, despite encountering some technical issues.
Last year, Alibaba was among the first to respond to DeepSeek's rapid expansion by releasing Qwen 2.5-Max, a model it claimed to be superior to one of DeepSeek's popular offerings.
Alibaba's most recent announcement did not mention DeepSeek. The comparative benchmarks provided focused on Qwen 3.5 against its own earlier versions and prominent US models such as GPT-5.2, Claude Opus 4.5, and Gemini 3 Pro.
DeepSeek is anticipated to launch its next-generation AI model in the near future, a development that is being closely monitored by investors and industry professionals following the company's significant impact on the global tech stock market last year.
Comments (0)
You must be logged in to comment.
Be the first to comment on this article!