Baidu Group Released Its First-Quarter 2025 Financial Report on the Evening of May 21, Revealing a Strong Performance. The report shows that Baidu achieved operating revenue of 32.5 billion yuan in the first quarter of 2025, with core revenue reaching 25.5 billion yuan, representing a year-on-year growth of 7% and surpassing market expectations. In the first quarter of 2025, Baidu's core net profit amounted to 7.63 billion yuan, marking a year-on-year increase of 48%, while its smart cloud business continued to grow, with a year-on-year increase of 42%.
Prof. Zhang Guangli from CUFE Business School shared his insights in an interview, stating that Baidu's first-quarter 2025 financial report highlights two strategic cores: accelerating commercialization driven by AI technology and deepening its global ecological layout. From the group's first-quarter report, it is evident that behind the stable financial growth lies the strategic transformation and upgrading of Baidu Group.
Seizing Long-Term Growth Opportunities in the AI Era
Baidu's founder, Robin Li, stated, "Our smart cloud business has shown strong performance, highlighting the market's growing recognition of our unique advantage in providing cost-effective full-stack AI products and solutions."
"We believe that our AI-first strategy enables us to maintain a leading position and seize long-term growth opportunities in the AI era," Li said.
Shanghai Securities News reporters found that Baidu has recently launched multiple Wenxin large models. In March this year, Baidu's first multimodal foundation model, Wenxin 4.5, and its first reasoning model, Wenxin X1, were officially launched. In April, Baidu introduced the enhanced Wenxin 4.5 Turbo and Wenxin X1 Turbo.
Baidu Group Vice President Wu Tian stated that Wenxin Large Model 4.5 is a multimodal foundation model, and Wenxin 4.5 Turbo, derived from Wenxin 4.5, offers better performance at a lower cost. "Additionally, based on Wenxin 4.5 Turbo, Wenxin X1 has been upgraded to X1 Turbo. While improving performance, it now features an advanced chain of thought, further enhancing capabilities in question answering, creation, logical reasoning, tool invocation, and multimodal functions," Wu said.
It is reported that the price of Wenxin 4.5 Turbo has decreased by 80% compared to Wenxin 4.5, with the input price per million tokens at 0.8 yuan and the output price at 3.2 yuan. The Wenxin Large Model X1 Turbo features an advanced chain of thought, further enhancing capabilities in question answering, creation, logical reasoning, tool invocation, and multimodal functions. Compared to Wenxin X1, its performance has improved while the price has decreased by 50%, with the input price per million tokens at 1 yuan and the output price at 4 yuan.
During the reporting period, Baidu Smart Cloud's Qianfan Large Model Platform was upgraded, expanding the model library and offering a more comprehensive toolchain to support training and fine-tuning of multimodal and deep-thinking models, further promoting the development of AI applications.
In Zhang Guangli's view, smart cloud serves as Baidu's second growth curve and has become a powerful profit engine for Baidu Group. Baidu's AI foundation continues to iterate and upgrade, AI application scenarios are continuously expanding, and the profit model is essentially established. This validates the synergy between technical investment and commercial efficiency under the "AI-first" strategy, laying the foundation for long-term growth.
Baidu Accelerates AI and Global Layout
In terms of autonomous driving, as of May 2025, Apollo Go has provided over 11 million global travel services. In the first quarter of this year, Apollo Go provided over 1.4 million global travel services, representing a year-on-year increase of 75%.
"Our autonomous driving travel services have also made significant progress. Apollo Go has expanded into global markets, entering Dubai and Abu Dhabi, demonstrating our commitment to providing safe, comfortable, and convenient autonomous driving travel services for global markets," Robin Li said.
Baidu stated that in the first quarter, Apollo Go accelerated its globalization, entering Dubai and Abu Dhabi. Currently, Apollo Go has initiated public road validation tests in Dubai, and the testing area in Hong Kong has also been further expanded. Additionally, Apollo has formed a strategic partnership with Shenzhou Car Rental to jointly create the world's first autonomous driving car rental service.
"Baidu is accelerating its overseas layout for Apollo Go, and its international strategy has achieved preliminary breakthroughs," said Prof. Qu Qixing from the School of International Economic and Trade at the University of International Business and Economics. He further analyzed that the commercialization path of autonomous driving relies far more on ecological collaboration and policy environment adaptation than on technological development. The sustainability of its overseas expansion will largely depend on Baidu's ability to integrate resources, ensure compliance, and match business models in overseas markets.
The financial report shows that in March this year, the monthly active users of the Baidu App reached 742 million, representing a year-on-year growth of 7%. On April 28, 2025, Baidu released its annual ESG report. The report indicates that as of December 2024, Baidu has trained 5.92 million AI talents.
Recently, Robin Li announced, "Over the next five years, Baidu will ramp up efforts to train an additional 10 million AI talents for society."
The report shows that Baidu's MSCI ESG rating has been upgraded to A, and it has been selected for the S&P Global Sustainability Yearbook 2025. It is reported that in the pan-internet technology industry, only 11 representative companies globally made the list, including Alphabet, Google's parent company, and Tencent.
It is understood that Baidu's "5 Million AI Talents Plan" proposed five years ago was completed ahead of schedule in April 2024. Building on the completion of the "5 Million AI Talents Plan," Baidu launched the "Baidu Talent Training Galaxy Plan," aiming to train an additional 5 million large model talents.
