Alibaba Launches Qwen 3.5 AI Model to Rival OpenAI & Google

by mark.thompson business editor

The competition to build the next generation of artificial intelligence is intensifying, and Alibaba is making a significant move. On Monday, the Chinese tech giant unveiled Qwen3.5, its latest large language model, boasting improved performance, reduced costs, and a new focus on “agentic AI” – systems capable of independent action and complex task completion. The release comes as Chinese companies race to catch up with, and in some cases surpass, Western AI leaders like OpenAI, Anthropic, and Google DeepMind in the development of increasingly sophisticated AI tools.

Qwen3.5 is available in two versions: an open-weight model allowing users to download, customize, and deploy the AI on their own infrastructure, and a hosted version accessible through Alibaba’s cloud platform, Model Studio. This dual approach caters to a wide range of users, from researchers and developers to businesses seeking to integrate AI into their operations. The timing of the release, coinciding with the eve of the Chinese New Year, is notable, signaling Alibaba’s commitment to staying at the forefront of this rapidly evolving field. This launch follows Alibaba’s release last week of an AI model specifically designed for robotics, demonstrating a broad strategy across multiple AI applications.

Sopa Images | Lightrocket | Getty Images

A Shift Towards AI Agents

The key differentiator for Qwen3.5 is its support for “agentic capabilities.” According to Alibaba, these capabilities allow the model to not just process information, but to independently take actions and complete multi-step tasks with minimal human oversight. This is a major trend in the AI world, spurred by recent advancements from American companies like Anthropic, whose new agent tools have garnered significant attention. The potential impact of these AI agents is substantial, with some analysts suggesting they could disrupt the software-as-a-service industry and beyond.

Qwen3.5’s compatibility with open-source AI agents, such as those from OpenClaw, further enhances its appeal to developers and researchers. OpenClaw, which recently experienced a surge in popularity, provides a framework for building and deploying AI agents, and Qwen3.5’s integration with this platform streamlines the development process. The creator of OpenClaw, Peter Steinberger, is joining OpenAI, according to a report from CNBC, signaling the growing importance of agentic AI within the industry.

Technical Specifications and Performance

The Qwen3.5 open-weight model features 397 billion parameters, which are the variables that determine how an AI system learns and reasons. While this is fewer parameters than some of Alibaba’s previous flagship models, the company claims significant performance improvements based on internal benchmark evaluations. Alibaba has released benchmark tests suggesting Qwen3.5 performs on par with leading models from OpenAI, Anthropic, and Google DeepMind. however, these comparisons are self-reported and have not been independently verified by CNBC.

The new model also boasts “native multimodal capabilities,” meaning it can understand and process text, images, and video simultaneously. This allows for more nuanced and comprehensive AI interactions. Qwen3.5 supports 201 languages and dialects, a substantial increase from the 82 supported by the previous generation, expanding its potential global reach. Alibaba plans to release additional open-weight models throughout the Chinese New Year period, according to Lin Junyang, technical lead of Alibaba Cloud’s Qwen team, in a social media post.

China’s AI Race Heats Up

Alibaba isn’t alone in pushing the boundaries of AI in China. Competitors like ByteDance and Zhipu AI have also released upgraded models in the past week, all focused on enhancing agentic capabilities. This flurry of activity underscores the intense competition within the Chinese AI landscape. Google DeepMind head Demis Hassabis told CNBC last month that Chinese AI models were “just months” behind Western rivals, a claim that suggests the gap is rapidly closing.

The development of Qwen3.5 and similar models is part of a broader effort by China to become a global leader in artificial intelligence. The Chinese government has invested heavily in AI research and development, and companies like Alibaba are playing a key role in driving innovation. The focus on AI agents represents a strategic shift, aiming to create AI systems that are not just intelligent, but also proactive and capable of solving real-world problems.

The release of Qwen3.5 marks a significant step forward in China’s AI ambitions. As the technology continues to evolve, it will be crucial to monitor the progress of both Chinese and Western companies, and to understand the implications of these advancements for businesses, governments, and society as a whole. The next step for Alibaba will be the continued release of open-weight models throughout the Chinese New Year, as announced by Lin Junyang, and further refinement of the Qwen3.5 platform based on user feedback and real-world applications.

Stay tuned to time.news for ongoing coverage of the rapidly evolving world of artificial intelligence.

You may also like

Leave a Comment