On Thursday, Chinese e-commerce juggernaut Alibaba Group Holding BABA launched a new reasoning model comparable to that of its affordable rival DeepSeek’s R1, Reuters reports.
Alibaba Qwen artificial intelligence unit tweeted that its QwQ-32B, with 32 billion parameters, can achieve performance comparable to DeepSeek’s R1 model.
On Wednesday, China pledged increased support for industries including AI, humanoid robots, and 6G telecom, as AI models noted growing adoption by government agencies and smaller companies in China.
Also Read: Taiwan Semiconductor Caught Between US Tariff Threats and Taiwan’s Push To Keep Advanced Chip Tech
Alibaba said the new model is accessible via its chatbot service, Qwen Chat. Users can choose various Qwen models and demonstrate mathematical reasoning, coding, and general problem-solving capabilities in benchmark tests.
Analysts told Reuters that government initiatives will expand usage of AI in China.
Beijing plans to issue 1.3 trillion yuan ($179 billion) in ultra-long special treasury bonds in 2025, up from 1 trillion yuan in 2024. It will allow local governments to issue 4.4 trillion yuan in special debt. The budget deficit will likely widen to around 4% of GDP.
Alibaba was among the first major global tech companies to open-source its large-scale AI models.
Alibaba committed $53 billion to cloud and AI infrastructure over the next three years.
DeepSeek’s AI model price war offers developers discounted off-peak pricing. Chinese tech companies like Alibaba and its rivals were already engaged in a price war to win market share.
Benchmark’s Fawne Jiang hailed Alibaba as a leading player in China’s AI sector, backed by its edge in infrastructure, proprietary models, and application capacity across diverse use cases.
Price Action: BABA stock closed higher by 8.61% at $141.03 on Wednesday.
Also Read:
Photo by jackpress via Shutterstock
© 2025 Benzinga.com. Benzinga does not provide investment advice. All rights reserved.