The company says its open-source model QwQ-32B is best at mathematics and coding.

Alibaba has launched a new artificial intelligence (AI) model, QwQ-32B, which it claims is smaller in size but stronger in performance. The model, which specializes in reasoning, mathematics, and coding, has been made open-source, allowing developers worldwide to access and use it freely for commercial purposes.

The release of QwQ-32B marks a significant step in Alibaba's AI ambitions and almost coincided with the launch of Manus, the creation of Chinese company Monica.

The Alibaba model has undergone extensive reinforcement learning, making it highly efficient in problem-solving. Benchmark tests show that QwQ-32B outperforms OpenAI’s o1-mini model and competes closely with DeepSeek-R1, another Chinese open-source AI model.

More to read:
Investigation: Moscow has learned to manipulate Western AI chatbots

One of QwQ-32B’s key advantages is its ability to operate on consumer-grade hardware, reducing deployment costs. This makes it an attractive option for businesses and developers seeking powerful AI capabilities without investing in expensive infrastructure.

It also integrates intelligent agent functionalities, enabling the model to adapt its reasoning based on real-time feedback.

Alibaba’s move comes at a time of intense competition in the AI industry. DeepSeek’s R1 model, released earlier this year, surprised the market with its high performance and cost efficiency. Alibaba has responded by not only launching QwQ-32B but also committing over $52 billion to AI and cloud computing development over the next three years.

The announcement has had a positive impact on Alibaba’s market position, with its Hong Kong-listed shares rising by 8% following the news.

More to read:
Chinese EV makers rush to integrate DeepSeek AI amid fierce market competition

QwQ-32B is now available for free download on platforms like Moda Community and Hugging Face. Developers can deploy it locally or integrate it into cloud-based applications using Alibaba Cloud services.

While QwQ-32B reviews are scarce, some tech-savvies tested Manus. Although it’s billed as "a general AI agent that turns your thoughts into actions," Axios said data security and privacy are of big concerns at this model while TechCrunch found its performance on various tasks “disappointing” – with error messages, endless loops and mistakes on factual questions.

It remains to be seen whether Alibaba’s QwQ-32B performs any better.

The rise of Chinese AI applications is a reason for concern in the U.S. tech industry, which has invested heavily in development of large language models while generating modest revenues.

***
NewsCafe is an independent outlet. Our sources of income amount to ads and subscriptions. You can support us via PayPal: office[at]rudeana.com or https://paypal.me/newscafeeu, or https://buymeacoffee.com/newscafe - any amount is welcome. You may also want to like or share our story, that would help us too.