Alibaba debuts Qwen3.5 model for agentic AI push
Chinese technology giant Alibaba on Monday unveiled a new artificial intelligence model, Qwen3.5, designed for what it describes as the emerging “agentic AI era,” highlighting major gains in performance and cost efficiency that it says outperform leading U.S. competitors on several benchmarks.
The launch is part of Alibaba’s broader push to expand the user base of its Qwen chatbot application in China’s increasingly competitive AI landscape, News.Az reports, citing Reuters.
The market is currently dominated by ByteDance’s Doubao and rising AI firm DeepSeek, which became the first Chinese artificial intelligence company to gain significant global traction last year.
Alibaba said Qwen3.5 is 60% cheaper to operate than its predecessor and eight times more capable of handling large workloads. The company also emphasized the model’s ability to independently carry out actions across mobile and desktop applications — a feature it calls “visual agentic capabilities.”
“Built for the agentic AI era, Qwen3.5 is designed to help developers and enterprises move faster and do more with the same compute, setting a new benchmark for capability per unit of inference cost,” Alibaba said in a statement.
The announcement follows ByteDance’s release of Doubao 2.0 on Saturday. The upgraded chatbot, which reportedly commands nearly 200 million users in China, was also marketed as optimized for the AI agent era, underscoring intensifying competition among major Chinese technology firms.
The rollout of Qwen3.5 could build on recent momentum for Alibaba in the domestic AI race. Earlier this month, the company ran a coupon campaign encouraging consumers to purchase food and beverages directly through the Qwen chatbot. The promotion led to a seven-fold increase in active users, despite reports of occasional technical glitches.
Alibaba was among the first companies to respond to DeepSeek’s rapid rise last year, releasing Qwen 2.5-Max, which it claimed outperformed one of DeepSeek’s flagship models.
In its Qwen3.5 announcement, Alibaba did not mention DeepSeek directly. The performance benchmarks published by the company show the new model surpassing its earlier iteration as well as prominent U.S. models, including GPT-5.2, Claude Opus 4.5, and Gemini 3 Pro.
DeepSeek is widely expected to unveil its next-generation model in the coming days. Anticipation remains high among investors and industry observers, particularly after the company’s breakthrough last year triggered a sharp global selloff in technology shares.
News.Az
By Nijat Babayev
