百度Q1财报：AI收入首超在线营销 广告占比跌破50%

The most notable takeaway from Baidu's Q1 2026 earnings report is that its AI business revenue has surpassed its traditional online marketing business for the first time, marking Baidu's official entry into a new AI-driven growth phase. Key highlights are as follows:

1. Core performance metrics: Total revenue reached 32.1 billion yuan, down 2% year-over-year. Net profit from core operations hit 3.4 billion yuan, and total cash reserves stand at 279.3 billion yuan. AI revenue came in at 13.6 billion yuan, accounting for 52% of core operating revenue. Meanwhile, online marketing revenue dropped 22% year-over-year to 12.6 billion yuan, falling to 48% of core revenue and dropping below 50% for the first time.

2. Baidu's AI development strategy: Baidu is not chasing larger large language model (LLM) parameter sizes or blindly burning cash to scale up. Instead, it focuses on four core application verticals, iterating its LLM capabilities around concrete product requirements.

3. Industry commercialization outlook: In the long run, AI services will shift to outcome-based pricing rather than token-based consumption billing. The future AI application market will be far larger than the current computing power market.

Baidu's latest earnings reveals new shifts and opportunities for brands in the AI era. Key takeaways for brands are as follows:

1. New marketing opportunities: Baidu's AI digital humans now deliver conversion rates comparable to human streamers in e-commerce live streaming, with the added advantages of hyper-realism and low cost. Brands can test AI digital human live streaming to cut customer acquisition costs. Its Fame Strategist enterprise agent also helps brands optimize decision-making in complex, dynamic environments and improve operational efficiency.

2. Product R&D guidance: AI has now entered the implementation phase. Brands should not blindly chase LLM parameter size trends when developing AI-related projects; instead, they should align AI capabilities with their own business and real user needs.

3. Cost and model reference: AI services will eventually shift to value-based pricing. Brands can focus on actual implementation outcomes rather than raw computing power consumption when purchasing AI services, making it easier to control costs. They can also use Baidu's Miaoda low-code platform to quickly build custom AI applications tailored to their specific business needs.

Baidu's Q1 2026 earnings highlights new growth opportunities and risk points in the AI sector. Key insights for sellers are as follows:

1. New incremental market opportunities: Demand for AI infrastructure is booming, with Baidu's GPU cloud business growing 184% year-over-year, driving rapid expansion of the entire AI computing service market. Sellers of related supporting services can capitalize on this growth wave. AI digital humans have already proven their conversion effectiveness in live streaming, so sellers can build out AI live streaming channels at low cost to expand sales scenarios. Autonomous driving has already launched in multiple overseas cities, so cross-border sellers can also explore related supporting partnership opportunities.

2. Risk warning: The AI application layer is still in an early development stage and has not yet generated scaled revenue. Sellers should avoid making large blind investments in AI application projects, which could lead to heavy spending without returns.

3. Recommended business models: Sellers can adopt differentiated pricing for different products, combining usage-based billing, subscriptions and other models, and center profit model design around delivering tangible user value.

Baidu's AI transformation experience and the industry shifts revealed in its earnings offer multiple insights for factories advancing digital transformation and unlocking new business opportunities. Key takeaways are as follows:

1. Guidance for digital transformation direction: AI technology has now entered the implementation phase. Factories do not need to blindly pursue LLMs with top-tier parameter sizes or follow industry trends when advancing digitalization. Instead, they should align AI capabilities with the concrete real-world needs of their production, design and operations workflows, and prioritize implementation outcomes.

2. Accessible low-cost AI tools: Baidu's Miaoda low-code platform allows users to build applications via natural language. Factories do not need to hire high-salaried senior technical talent to quickly build digital management and design tools tailored to their own needs, drastically lowering the barrier to digital transformation.

3. New business opportunities: Demand for AI computing infrastructure such as GPUs is growing explosively. Factories producing related supporting hardware can capitalize on the current wave of AI infrastructure buildout to expand business growth. Demand for digital humans is also rising rapidly, so factories in the related supporting supply chain can tap into new demand growth points.

Baidu's latest earnings reveals key development trends, customer pain points and actionable solutions for the AI service industry. Key insights are as follows:

1. Industry development trends: AI commercialization is still dominated by computing infrastructure output. Demand for GPU cloud is growing explosively, with 184% year-over-year growth, and it inherently delivers higher margins than traditional CPU cloud services, making it the core growth track for AI service providers today. The AI application layer is still in early development but offers enormous long-term growth potential; the future AI application market will be far larger than the computing layer market.

2. Core customer pain points: Customers now prioritize the actual value AI services deliver over the amount of computing power consumed. The traditional token-based billing model cannot align with customers' long-term needs, so the industry's pricing logic must evolve.

3. Recommended development strategies: Service providers can follow Baidu's example: avoid blindly competing on LLM parameter sizes to climb leaderboards, and instead focus on refining product capabilities for specific application scenarios, iterating technology around real customer needs. At this stage, providers can adopt differentiated pricing models for different products, and gradually explore commercialization paths that fit their business.

Baidu's AI transformation practice and the industry shifts revealed in its earnings offer key insights for platform operators. Key takeaways are as follows:

1. Core market demand for AI platforms: Both enterprises and end users prioritize AI applications that solve practical problems over competition in LLM parameter sizes. Platforms should adjust their ecosystem guidance to encourage developers to focus on scenario-based implementation, rather than purely competing on technical benchmarks.

2. High-priority growth areas: Demand for GPU cloud is growing rapidly, customers prioritize service stability, and GPU cloud delivers higher margins than traditional businesses. Platforms can expand their GPU cloud布局 to capture current incremental demand while optimizing their overall profit structure.

3. Trend guidance and risk mitigation: The AI application layer is still in the early stage of commercialization and has not yet generated scaled revenue, and capital markets have limited patience for slow-monetization applications. Platforms should not blindly burn cash to chase AI application trends or scale at the cost of heavy losses, and should instead control the pace of investment, expanding gradually only after validating business models. Platforms can adopt differentiated pricing for different scenario-based products, centering profit model design around user value.

Baidu's Q1 2026 earnings marks that the AI industry has entered a brand-new development stage, with multiple notable new industry trends and open research questions. Key insights are as follows:

1. New industry trends: Chinese tech giant Baidu has achieved a replacement of traditional online marketing business by its AI business, making AI officially its first growth curve. AI industrialization has first achieved scaled profitability at the computing layer, while the application layer remains in early exploratory stage. Industry competition logic has already shifted from competing on LLM parameter sizes to competing on application implementation capabilities. Leading companies have already started adjusting organizational structures to align R&D resources with application scenario requirements.

2. New business model exploration: The current mainstream pricing model in the AI industry is token-based billing, but this will shift to outcome-based pricing in the long run. The agent-based service market will eventually become a much larger new market than the current market, and differentiated pricing is a reasonable choice for the current exploratory stage.

3. Open research questions: The long-term impact of the continuous decline of traditional online marketing business and slowing monthly active user growth during the transition period has not yet emerged. It remains unclear whether capital market patience will be sufficient to support the industry through the exploration phase, given that AI application layer monetization is slower than expected. These are all new industry-level questions that require in-depth research. Baidu's autonomous driving overseas expansion also offers a new research case for Chinese AI globalization.