In a bold statement of China’s growing influence in the worldwide AI race, Alibaba has unveiled Qwen3, a new class of state-of-the-art artificial intelligence models, in a capable move that is crafted to match and in some cases even exceed those of industry giants such as OpenAI, Google, and Anthropic. The Qwen3 series is Alibaba’s latest step in its long-term investment in AGI, representing a quantum leap in both AI scale and performance and accessibility.
Constructed with different models ranging from 0.6 billion to 235 billion parameters, Qwen3 provides both low and high performance-efficient models for a wide range of applications. An innovative hybrid reasoning architecture makes Qwen3 stand out, as it switches between quick-response and slow-thinking modes. This means the model can react fast to easy problems but use a methodical approach when handling complex tasks—a two-mode method that’s highly reminiscent of the way humans think.
Users can even control the model’s “thinking budget,” tuning performance depending on trade-offs for a particular task. Alibaba’s decision to open-source the complete Qwen3 series—now available through platforms including Hugging Face and GitHub—is part of a wave of industry attraction to transparency and collaboration. This strategy will also place Qwen3 as a direct competitor of other open models like DeepSeek and the Claude series of Anthropic, which are already popular for delivering high-performing pre-trained models at lower compute costs.
Two of the models in the Qwen3 family employ a mixture-of-experts (MoE) design, which splits intricate jobs into smaller tasks that are performed by dedicated “experts” in a neural network. This approach improves efficiency and makes the models extremely scalable, a feature that will be attractive for developers interested in cost-effective deployment on cloud, desktop, and even mobile.
Besides outperforming benchmarks, Qwen3—like its predecessors Qwen and Qwen-X—is impressively capable, given that it is trained on 36 trillion tokens, a dataset comprising textbooks, code, Q&A samples, and AI-generated content. Its polyglot and polyformat fluency will indeed further cement Alibaba’s competitive position in the race to democratize AI.
The release is hitting at a time when China’s artificial intelligence ecosystem is bursting at the seams. Spurred by what they see as DeepSeek’s game-changing trajectory, Chinese tech firms have been turning out new models at a frenetic pace. In 2025, Alibaba announced it was “all-in” on AI and had very recently launched Qwen 2.5: a multimodal system for text, images, audio, and video, efficient enough to run on a personal device.
This week’s release of Qwen3 is more than a technical upgrade; it’s a strategic statement that Alibaba is going to build its AI future around the Qwen brand. As competition with these Western AI labs continues to escalate, the company’s open-source approach and hybrid reasoning technology provide an interesting option in the race for AI dominance worldwide.
On April 28, 2025, Alibaba Group Holding Ltd. announced Qwen3, a new family of artificial intelligence models that it says can outperform competing models from companies such as OpenAI and Google LLC. The new release underscores the rapid pace of development within China’s AI industry since DeepSeek Ltd. first burst onto the scene late last year. The e-commerce giant said the new Qwen3 models surpass the capabilities of DeepSeek’s best models in several areas, including coding and math-based problems.
Alibaba is releasing a number of models within the Qwen3 family under an open-source license, ranging in size from 600 million to 235 billion parameters, which is a measure that roughly corresponds to problem-solving abilities. As a rule, the more parameters a model has, the better it performs. Within the new Qwen3 series are two “mixture-of-experts” or MoE models that Alibaba says are able to compete with the most advanced reasoning models launched by Google and Anthropic PBC.
Reasoning models are designed to mimic the way humans think about problems, taking more time to consider things and perform fact-checking for accuracy. By using the MoE technique, AI models can enhance their reasoning skills by dividing a task into smaller segments, similar to how a company might employ teams of specialists to focus on specific parts of a more challenging problem.
“We have seamlessly integrated thinking and non-thinking modes, offering users the flexibility to control the thinking budget,” Alibaba’s Qwen team said in a blog post. “This design enables users to configure task-specific budgets with greater ease.” Alibaba said the Qwen3 models support 119 languages and have been trained on a dataset containing almost 36 trillion tokens, which are the raw bits of data they process while being “taught.” One million tokens is equivalent to around 750,000 words, and in this case, the data was drawn from various textbooks, code snippets, AI-generated data, question-answer pairs, and other sources.
In various benchmark tests, Alibaba’s Qwen3 models delivered impressive results, edging out recent “high-end” models from U.S.-based AI companies, such as OpenAI’s o3-mini and o4-mini models. For instance, on the Codeforces benchmark that measures models’ ability to write code, the largest Qwen-3-235B-A22B model surpassed o3-mini and Google’s Gemini 2.5 Pro. It also beat o3-mini on the AIME mathematics benchmark, as well as the BFCL test that assesses AI models’ reasoning abilities.
Since DeepSeek’s R1 reasoning model first burst onto the scene at the end of December, upstaging OpenAI despite being developed at just a fraction of the cost, Chinese tech leaders have released a flurry of similarly powerful AI models. Alibaba launched the Qwen-2.5 series models just a few weeks earlier, noting they can process multimodal data formats including text, images, audio, and video. Those models are notably lightweight, designed to run directly on smartphones and laptops.
The emergence of powerful, open-source Chinese AI models has upped the ante for U.S. AI companies, which were until recently seen as industry leaders. But their status is coming under threat, especially because American-made models are generally trained at much higher costs than their Chinese counterparts. The U.S. government has also responded, introducing further sanctions that aim to prevent Chinese companies from getting their hands on the powerful graphics processing units used to train and run AI models.
In its most recent move, the U.S. slapped an export license on Nvidia Corp.’s H20 GPU, which had been designed specifically to comply with earlier sanctions on China. OpenAI has responded by saying it will release an “open-weights” reasoning model in the next few months, marking a dramatic reversal from its usual approach, where the inner workings of its models are essentially a “black box.”
Alibaba Chief Executive Officer Eddie Wu said in February that the company’s main objective is to build an “artificial general intelligence” system that will ultimately surpass the intellectual capabilities of humans. Qwen3 represents a thoughtful evolution in large language model development, addressing many of the core challenges that continue to affect LLM deployment today. Its design emphasizes adaptability—making it equally suitable for academic research, enterprise solutions, and future multimodal applications.
Looking ahead, the Qwen team positions Qwen3 not just as an incremental improvement but as a significant step toward future goals in Artificial General Intelligence (AGI) and Artificial Superintelligence (ASI), AI significantly smarter than humans. Plans for Qwen’s next phase include scaling data and model size further, extending context lengths, broadening modality support, and enhancing reinforcement learning with environmental feedback mechanisms.
