Baidu, Inc. Q1 FY2025 earnings call

Key Points

• AI Cloud: Revenue reached RMB6.7 billion in Q1, up 42% YOY, with non-GAAP operating profit remaining positive. Gen AI and foundation model related revenue saw triple-digit YOY growth. Qianfan platform was upgraded with expanded model library and enhanced toolkits for training, fine-tuning, and distillation of models.
• Baidu Search: Accelerated AI transformation, with 35% of mobile search result pages containing AI-generated content in April, up from 22% in January. Focus on multimodal content to enhance user experience.
• Intelligent Driving: Apollo Go achieved 100% fully driverless operations in Mainland China and expanded to Dubai and Abu Dhabi, with over 1,000 fully driverless vehicles deployed globally.
• Applications: Officially launched MiaoDA, a no-code capability tool, to democratize AI application development.

Baidu Core's total revenue in the first quarter reached RMB25.5 billion, a 7% year-over-year increase. The AI Cloud business was a key driver, with AI Cloud revenue reaching RMB6.7 billion in Q1, up 42% year-over-year, accounting for 26% of Baidu Core's revenue. Baidu Core's online marketing revenue was RMB16.0 billion, down 6% year-over-year, while non-online marketing revenue was RMB9.4 billion, up 40% year-over-year, primarily driven by the AI Cloud business.

Forward-Looking Statements

• Continue to invest in AI, including AI Cloud infrastructure, model iterations (e.g., ERNIE 5.0), and autonomous driving.
• Expect AI Cloud revenue to continue growing as AI adoption accelerates across sectors.
• Plan to further expand global presence of Apollo Go and enhance monetization in AI search over the long term.

Risks

• Forward-looking statements subject to risks and uncertainties that may cause actual results to differ from expectations, including risks discussed in Baidu's Annual Report and SEC/Hong Kong Stock Exchange filings.
• Potential impacts from factors like U.S. export restrictions on advanced AI chips, though Baidu's full-stack AI capabilities aim to mitigate such impacts.

Q: Please share the latest update on AI overall strategy and technology roadmap, including ERNIE 5.0 and inference cost reduction.

A: Robin Li stated Baidu sticks to an application-driven approach, focusing model iteration on areas with real application value. Working on next-gen models, and each new model has seen significant price cuts. Set to open source ERNIE 4.5 series on June 30.

Q: What are the key drivers for AI Cloud growth and outlook for profitability?

A: Dou Shen mentioned strong demand for Gen AI and foundation models across industries. AI Cloud revenue breakdown includes enterprise cloud with subscription-based revenue growing, and non-GAAP operating margin expanding. AI chip export restrictions are mitigated by Baidu's full-stack capabilities.

Q: Rationale behind accelerating AI search transition, expectations for AI answers penetration, and monetization updates.

A: Robin Li said AI search transition is to enhance user experience. April saw 35% of mobile search results as AI-generated content, expected to rise in Q2. Monetization for AI search is in early stages but sees long-term potential to monetize more queries.

Q: Evolving competitive landscape for robotaxi, differentiation of RT6, and expansion plans.

A: Robin Li said RT6 is world's first purpose-built Level 4 autonomous vehicle, with low unit cost. Apollo Go has 15 cities globally, expanding to Dubai and Abu Dhabi, and exploring partnerships for faster expansion.

Q: How Baidu competes with other AI applications and platforms?

A: Robin Li mentioned Baidu was early to use AI to transform legacy core business. Focus on AI search to efficiently meet user needs with multimodal content and reimagining search to go beyond information finding using agents.