Understanding the Market | Baidu Group-SW rose over 3% in early trading, officially releasing the Wenxin large model X1.1, with overall performance surpassing DeepSeek R1-0528
Baidu Group-SW rose over 3% in early trading, and as of the time of writing, it is up 3.3%, priced at HKD 109.6, with a transaction volume of HKD 702 million. In terms of news, on September 9th, according to Baidu's official Weibo account, the Wenxin large model X1.1 was officially released at the WAVE SUMMIT Deep Learning Developer Conference 2025. It was introduced that users can currently use the Wenxin large model X1.1 on the Wenxin Yiyan official website and the Wenxiao Yan APP. The Wenxin large model X1.1 has also officially launched on Baidu Intelligent Cloud's Qianfan platform, fully open for enterprise clients and developers. Compared to the Wenxin large model X1, X1.1 has improved factuality by 34.8%, instruction adherence by 12.5%, and agent performance by 9.6%. Additionally, in several authoritative benchmark evaluations, the overall performance of the Wenxin large model X1.1 surpassed DeepSeek R1-0528, demonstrating a leading advantage in certain tasks. At the same time, when compared to international top models GPT-5 and Gemini 2.5 Pro, its performance is on par
According to Zhitong Finance APP, Baidu Group-SW (09888) rose over 3% in early trading, and as of the time of writing, it is up 3.3%, priced at HKD 109.6, with a transaction volume of HKD 702 million.
In terms of news, on September 9th, according to Baidu's official Weibo account, the Wenxin large model X1.1 was officially released at the WAVE SUMMIT Deep Learning Developer Conference 2025. It was introduced that users can currently use the Wenxin large model X1.1 on the Wenxin Yiyan official website and the Wenxiao Yiyan APP. The Wenxin large model X1.1 has also officially launched on Baidu Intelligent Cloud's Qianfan platform, fully open for enterprise customers and developers. Compared to the Wenxin large model X1, X1.1 has improved factuality by 34.8%, instruction adherence by 12.5%, and agent performance by 9.6%. Additionally, in several authoritative benchmark evaluations, the overall performance of the Wenxin large model X1.1 surpassed DeepSeek R1-0528, demonstrating a leading advantage in certain tasks. At the same time, its performance is on par with international top models GPT-5 and Gemini 2.5 Pro
