Alibaba Sets Sights on ChatGPT and Gemini With New AI Model

Published: Jan. 27, 2026 5:49 p.m. GMT+8

00:00

00:00/00:00

Listen to this article 1x

Alibaba released its new Qwen3-Max-Thinking model on Jan. 26. Photo: VCG

Alibaba Group Holding Ltd. on Monday unveiled its latest artificial intelligence (AI) model which it claimed outperforms rivals like OpenAI’s GPT-5.2 and Google’s Gemini 3 Pro across multiple benchmarks, intensifying the global race for dominance in generative AI.

Qwen3-Max-Thinking, which is available to enterprise users for API calls and to individual users on a trial basis, adopts a mechanism, which allows the model to use prior inference results to improve the quality of its output, Alibaba said.

You've accessed an article available only to subscribers

Subscribe today for just $.99.

VIEW OPTIONS

Unlock exclusive discounts with a Caixin group subscription — ideal for teams and organizations.

Subscribe to both Caixin Global and The Wall Street Journal — for the price of one.

Share this article

Open WeChat and scan the QR code

Explore the story in 30 seconds

Alibaba unveiled Qwen3-Max-Thinking, an AI model claimed to outperform OpenAI’s GPT-5.2 and Google’s Gemini 3 Pro in benchmarks like scientific knowledge, mathematical reasoning, and coding.
The model is available to enterprise users by API and to individuals for trial, and integrates with the Qwen AI app offering various AI services.
Alibaba Cloud’s AI computing and storage capacities grew over five times and four times, respectively, in the past year.

AI generated, for reference only

Alibaba Group Holding Ltd.: Alibaba Group Holding Ltd. has released its new AI model, Qwen3-Max-Thinking, which it claims surpasses OpenAI's GPT-5.2 and Google's Gemini 3 Pro in several benchmarks. This model will be integrated with the Qwen AI app, offering various AI-powered services. Alibaba Cloud's CTO, Zhou Jingren, emphasized the need for rapid innovation in the competitive AI landscape.

Google: The article mentions Google in relation to its AI model, Gemini 3 Pro. Alibaba claims their new model, Qwen3-Max-Thinking, outperforms Google's Gemini 3 Pro in several benchmarks like scientific knowledge, math, coding, and search tool calling, though it trails in function calling. The article highlights the competitive landscape in generative AI, with companies like Google actively involved in model development and iteration.

Alibaba Cloud Intelligence Group: Alibaba Cloud Intelligence Group's CTO, Zhou Jingren, emphasizes the importance of rapid AI model iteration to stay competitive globally. The group has seen significant growth in AI computing power and storage capacity over the past year. They are responsible for the development and release of advanced AI models like Qwen3-Max-Thinking, which integrates with the Qwen AI app offering various consumer services.

OpenAI: OpenAI is a company that develops artificial intelligence. Its GPT-5.2 model was mentioned in the article as a rival to Alibaba's new AI model, Qwen3-Max-Thinking. While Alibaba's model outperforms GPT-5.2 in several benchmarks like scientific knowledge and coding, it trails GPT-5.2 in code debugging. OpenAI released GPT-5.2 in December, following GPT-5 in August, and its ChatGPT's global monthly active users rose by 6% between August and November.

Sensor Tower: Sensor Tower is a data analytics firm that tracks app usage. According to their data, the global monthly active users of ChatGPT and Gemini saw increases of 6% and 30% respectively, between August and November. This information helps illustrate the growth in user bases for major AI models.

AI generated, for reference only