Alibaba Group on February 16 introduced Qwen3.5, presenting it as a flagship upgrade tailored for what it calls the “agentic AI era.” The launch comes at a time when competition in advanced artificial intelligence models is intensifying both within China and globally.
The newly unveiled Qwen3.5 is designed to independently execute complex tasks, reflecting a broader industry shift toward AI systems capable of acting as autonomous agents. According to Alibaba, the model brings notable gains in both performance and cost efficiency. In multiple task evaluations, the company said the 3.5 series consistently performs on par with state-of-the-art leading models.
“Compared to the 3 series, these models show a leap forward in both pure-text and multimodal capabilities. This version is functionally equivalent to snapshotqwen3.5-plus-2026-02-15,” the company stated.
Alibaba confirmed in a blog post that it is open-sourcing Qwen3.5-397B-A17B, also referred to as Qwen3.5-Plus. The model can process text, images, and video inputs, and generate text outputs. It also supports visual agentic actions across both mobile and desktop applications, positioning it for real-world workflow automation.
The Qwen3.5-397B-A17B variant demonstrates strong capabilities across language understanding, reasoning, code generation, agentic workflows, image and video comprehension, and graphical user interface (GUI) interaction. The company said it rivals leading frontier models in versatility and performance.
Alibaba added that the new model delivers up to a 60% reduction in deployment expenses while matching the performance of the significantly larger Qwen3-Max model, which exceeds one trillion parameters. The company further noted that Qwen3.5 runs large workloads eight times faster than its predecessor.
“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,” the company said, according to Reuters.
In benchmark testing, Alibaba said Qwen3.5 outperformed its earlier models as well as certain U.S.-developed systems including GPT-5.2, Claude Opus 4.5, and Gemini 3 Pro.
The release comes amid heightened competition in China’s rapidly evolving AI landscape. Companies such as ByteDance, Baidu, and DeepSeek are all advancing their own large-scale models.
ByteDance recently launched Doubao 2.0, positioning it for the “agent era.” Its AI video-generation model Seedance 2.0 has drawn attention overseas on platforms such as X, including commentary from Elon Musk. Meanwhile, DeepSeek, which gained prominence last year with its low-cost AI model, is expected to release a new version in the coming days.
Alibaba has also been actively promoting usage of its Qwen chatbot through ecosystem integration efforts. Despite encountering technical issues, a recent discount campaign aimed at encouraging chatbot adoption helped boost active user numbers this month. Last year, the company introduced Qwen 2.5-Max as one of the early major competitive responses to DeepSeek’s rapid rise.
The unveiling of Qwen3.5 underscores Alibaba’s continued emphasis on cost efficiency and functional breadth as enterprises and developers look for scalable AI solutions. As competition in the AI sector intensifies, market observers are likely to monitor user growth and monetization trends closely in upcoming earnings reports, particularly as companies race to define leadership in the emerging agentic AI landscape.
