Alibaba Group’s latest artificial intelligence (AI) development sent its Hong Kong-listed shares soaring over 8% on Thursday.
The surge followed the company’s announcement of its new AI reasoning model, QwQ-32B, which it claims rivals the widely recognized DeepSeek R1.
In a post on X, Alibaba’s AI division revealed that the QwQ-32B model, with 32 billion parameters, delivers performance comparable to DeepSeek’s R1, despite the latter boasting 671 billion parameters. The model is now accessible through Alibaba’s chatbot service, Qwen Chat, where users can choose from various Qwen models, including Qwen2.5-Max, the most advanced in the series.
Benchmark tests showed QwQ-32B excelling in mathematical reasoning, coding, and problem-solving, performing close to industry-leading models like OpenAI’s o1 mini and DeepSeek R1.
The announcement comes amid China’s growing commitment to AI development. On Wednesday, the Chinese government pledged increased support for emerging technologies, including artificial intelligence, humanoid robots, and 6G telecommunications.
Meanwhile, another AI release making waves on Thursday was Manus, an AI agent developed by Chinese startup Monica. The company claimed that Manus outperformed OpenAI’s Deep Research on a benchmark for AI assistants.
Unlike traditional chatbots, AI agents like Manus offer more advanced capabilities, assisting users with tasks such as travel planning and comparative analysis of insurance policies. Currently available by invitation only, a demonstration video posted on X late Wednesday attracted over 280,000 views, with many users requesting access.
