Author｜Chang Yuan

Editor｜Key Point

In 'The Nature of Technology,' technological thinker Brian Arthur offers a insightful judgment: the true power of a technology lies not in the moment of its invention, but when it begins to 'combinatorially evolve' and integrate into the existing economic structure.

Standing in mid-2026, AI large models are experiencing such a moment. When 'ERNIE Bot' and 'Tongyi Qianwen' are no longer confined to chat boxes but start operating your mouse and keyboard, taking over your workflow, and penetrating into bank risk control systems and power grid inspection routes, a battle over who can truly build an Agentic cloud has quietly begun.

After years of competition in the cloud computing market, no year has excited all players like 2026.

According to the latest IDC report, the market size of China's AI cloud full-stack services reached 28.09 billion yuan in the first half of 2025, a year-on-year increase of 195.7%. The demand generated by enterprises 'using AI' is real, but on the other hand, AI demand is reshaping the logic of industry cloud selection. The bidding market represents new trends in customer demand. In the first quarter of 2026, large model bidding projects were highly concentrated in industries such as finance, government affairs, energy, and transportation. Enterprises' core demands are no longer about renting hundreds of GPU cards for training but about AI truly solving business pain points. Banks want thousands of AI applications to run stably, automakers want assisted driving to be mass-produced and on the road, and energy companies want AI to enter front-end operations.

At the same time, policy is also fueling the trend. In May of this year, the Cyberspace Administration, National Development and Reform Commission, and Ministry of Industry and Information Technology jointly issued the 'Implementation Opinions on the Standardized Application and Innovative Development of Intelligent Agents.' AI is officially treated as industrial infrastructure, and the cloud is the core carrier of this vast 'intelligent transformation across all sectors.'

When customers demand not just rented resources but tangible results, the competitive dimension of cloud providers has changed. The past cloud competition was about who occupied more land; the new cloud war is about who takes deeper root. Bid amounts and customer numbers are just surface-level; the true determinant is penetration depth.

Therefore, the current competition among cloud business giants transcends the literal meaning of products and services. The real decisive battle in the 2026 AI cloud war is who can build an Agentic cloud that truly supports intelligent transformation across all sectors.

  What Are the Giants Competing for in Agentic Cloud?

Let's focus on specific players. Taking vendors that completed full-stack layout (layout) early as examples, Baidu and Alibaba have similarities but are also starkly different. Essentially, both are trying to transform AI from 'talkative' to 'capable.'

In the past few years, discussions centered on large models, with debates over parameters, rankings, and inference costs. But since this year, 'Agent' has become the absolute buzzword at various product launches. Alibaba talks about AI Native; Huawei proposes Pangu + Ascend + CloudMatrix; Baidu upgrades its original 'Cloud-AI Integration' to 'Chip-Cloud-Model-Agent.'

As IDC points out in its latest 'Global Cloud Computing Architecture Forecast 2026,' AI is reshaping the architecture and value proposition of cloud computing, with the cloud gradually transforming from a 'computing power pool' into a 'neural hub for AI operation and collaboration.' The competition among giants for Agentic cloud is no longer about 'speaking by scale' as in the traditional cloud market but is differentiated at the levels of AI capabilities and industry solutions.

Alibaba is betting on scale. From its self-developed chip HanGuang and Tongyi Qianwen to the BaiLian platform and the recently launched AI Native Agent platform, Alibaba is attempting to replicate its success path from the mobile internet era, building an ecosystem of 'Chip-Cloud-Model-Application.' According to Omdia, Alibaba Cloud has consistently maintained over one-third of China's cloud market share and 38.1% of the AI cloud market. The underlying logic of this approach is scale advantage. Like Taobao and Alipay in the past, it establishes a larger territory through ecosystem and investment, thereby driving more customer recognition.

In contrast, Baidu Intelligent Cloud has a similar approach in full-stack self-development and large-scale ecosystem construction but differs in application deployment and industry penetration—that is, in its definition of 'territory.'

As the earliest player in China to work backward from AI capabilities to cloud strength, Baidu began layout (laying out) deep learning in 2010, opened its cloud business in 2015, proposed 'Cloud-AI Integration' in 2020, and further upgraded to the new full-stack 'Chip-Cloud-Model-Agent' by 2026. Over more than a decade, Baidu's cloud business layout (layout) has not followed the traditional cloud vendor route of starting with resources, then intelligence, and scaling through resources. Instead, based on the judgment of 'building a cloud adapted to the Agent era,' it has made the endpoint itself the carrier of Agents.

This trend is evident from Baidu Intelligent Cloud's list of endpoint customers, which includes the global Top 10 smartphone manufacturers, China's Top 5 AI glasses companies, Top 5 AI toy manufacturers, Top 5 robotic vacuum brands, and even over 5 million smart home appliances. Behind these endpoints lies a vast imaginative space. IDC predicts that by 2028, over 1 billion Agents will run across different devices and business systems globally.

Agents will become the new digital workforce, and endpoints will be their largest carriers. In other words, every hardware device in the future may have its own brain. Future Agents are unlikely to be just an App but rather assistants in smartphones, guides in glasses, companions in toys, household managers in robotic vacuums, and even in every refrigerator, air conditioner, and washing machine. Baidu aims not just for a single super Agent but to connect the most scenarios.

This leads to a concept recently much discussed in the industry—FDE, Field Deployment Engineer, a field that Anthropic and OpenAI scrambled to layout (deploy) in early May and has become key to the last-mile delivery of AI. Baidu Intelligent Cloud has been working on industrial deployment for over a decade, from search and autonomous driving to energy and finance, accumulating possibly the largest FDE group in China. These individuals do not write papers or chase rankings but spend their days at customer sites, tuning models, aligning data, and fixing errors. When the industry began to realize the scarcity of FDEs, Baidu had already overcome the toughest phase.

An interesting divergence has emerged. In summary, Alibaba is building a bigger cloud, while Baidu is building a deeper cloud. The former competes on scale and ecosystem; the latter competes on industry and endpoints. In a sense, this is the biggest difference between Baidu Intelligent Cloud and other cloud providers.

  Success in Agentic Cloud Depends on Three Criteria

At any intelligent cloud conference this year, the questions from industrial CIOs and CTOs are all based on specific workflow details, down to 'Can the approval process be completed within three months?' or 'How to ensure no downtime or errors when integrating into core business systems within six months?' This on-site sense is unseen in the past 'resource-selling' cloud market. Take banks as an example: introducing an intelligent approval system requires handling millions of concurrent requests, private deployment within the bank, seamless integration with core host systems, and meeting security and compliance requirements—a complete service, not just computational power stacking but a systems integration project.

These demands have divided cloud service providers in the market into two camps: those still competing on resource scale and those capable of integrating AI capabilities into business processes. For cloud giants to build a true Agentic cloud, they must meet three hard criteria: chips, platforms, and industry penetration.

The first criterion: Are the chips self-developed and proven in real business?

In the Agent era, computational power is not just about being large but also stable, adaptable, and capable of long-term operation. The core logic is that if chips are unstable, Agents cannot run continuously, and enterprise operations will stall. Thus, the first threshold for an Agentic cloud is ensuring Agents can run stably in real environments over the long term. According to IDC's 'China AI Chip Market Report 2025,' no more than three vendors in China have an overall delivery stability rate exceeding 90% for enterprise-grade AI chips, and Baidu's Kunlun Core is one of them.

Today, the Kunlun Core P800 has completed large-scale validation of a 10,000-card cluster. In 2026, the Tianchi 256-card super-node completed ERNIE 5.1 large model training, improving throughput performance by 25% and inference efficiency by 50%, while supporting the on-demand construction of ultra-large clusters with hundreds of thousands or even millions of cards. In real business scenarios, the Kunlun Core has successfully run bank core processes and financial risk control. For example, China Merchants Bank deployed an AI chip cluster based on the P800, with over 800 AI applications running in production, of which over 50% run on domestic computational power.

The second criterion: Can the Agent platform withstand large-scale deployment?

After all, the true test for Agents is not running demos but operating long-term in error-intolerant scenarios like power grid inspections and bank risk control, handling millions of requests. If chips are the first hurdle for Agentic cloud deployment, Harness capability is the second. Simply put, when you have a smart brain (large model), how do you give it capable limbs and ensure it behaves? This system of 'adding limbs and setting rules' is Harness.

Especially in real industrial scenarios, an Agent faces problems a hundred times more complex than in labs to complete a task. Harness ensures the large model works efficiently and compliantly. Take power grid inspections as an example. When an inspection Agent receives the task 'Check equipment defects in this substation,' its workflow is not simple. It must first remember inspection standards (long-context management), then call the infrared camera's API (tool invocation). If one identification is uncertain, it must take another shot from a different angle (multi-step reasoning). Finally, it must compare the results with defect records from the past three months (persistent memory) to determine whether the defect is new or known. A single mistake in this process—whether calling the wrong camera, forgetting historical records, or getting stuck in reasoning—will cause missed detections.

This is the problem Harness solves. In its actual deployment, Baidu Intelligent Cloud breaks down this capability into several modules: long-context management, persistent memory, tool invocation, sub-agent scheduling, evaluation feedback, and Runtime environment. Through Harness Engineering at each stage, Agents remember tasks, operate tools, think step-by-step, and correct errors. This has reduced inspection time from 2.5 hours to 45 minutes and has been deployed in over 40 substation scenarios, covering more than 800 substations nationwide.

The third criterion: Has industry penetration formed a moat?

This is the most critical of the three criteria. Chip research and development progress is always a race, but penetration into core industry scenarios, once achieved, is hard to replace. For example, Shanghai Pudong Development Bank recently fine-tuned a financial analysis model with Baidu Intelligent Cloud in the core business process of corporate loan due diligence, compressing financial analysis from hours or even days to minutes, transforming expert experience into engineering and scalable capabilities. Not long ago, SPDB signed a strategic cooperation agreement with Baidu Group, deepening their partnership first established in 2020.

These customers are not piloting a single AI application but running AI in their main business processes or even specific scenarios. By extension, for industry intelligence, future 'scale' should not just refer to the number of industries entered but the specific scenarios where it plays a role and produces results.

If market share is the facade, penetration depth is the substance. Once formed, the latter becomes a moat. Baidu Intelligent Cloud's penetration data in several industries is also noteworthy. In the automotive industry, it covers 100% of China's mainstream automakers, supporting the delivery of over 20 million L2 assisted driving new cars in 2025. In the embodied intelligence sector, according to IDC, Baidu Intelligent Cloud leads with a 36.7% market share, exceeding the sum of the second and third players, with clients including leading companies like Beijing Humanoid and Unitree. Additionally, it partners with over 800 financial institutions, covering 100% of systemically important banks, and about 80 central enterprise clients in the energy sector. The significance of these numbers lies not just in breadth but also in depth.

Take finance as an example. According to IDC's latest report, even in the decision tools and services market for retail credit intelligent risk control, Baidu Intelligent Cloud maintains a leading market share, with its revenue and user base both doubling in the past year.

  The Decisive Point in 2026: From 'Bigger' to 'Deeper'

Over the past two years, large model vendors have been fiercely competing around Agents. The early popularity of OpenClaw confirmed that today, almost any company with basic model capabilities can build an Agent demo capable of calling tools, operating browsers, and completing complex tasks within weeks. However, the real challenge is getting these Agents into the real world and ensuring long-term stable operation.

After all, for banks, power grids, and automotive factories operating at high speeds daily, the usability and trustworthiness of Agents face stricter scrutiny in actual workflows. In banks, a single misjudgment can mean incorrect loan approvals; in automotive factories, a single failure can halt production lines; and a single grid failure can affect normal power supply for millions.

On the surface, everyone is building Agents, but if you zoom out, what several giants are competing for is the discourse power (discourse power) over the next generation of cloud infrastructure.

Looking at the past two decades of cloud computing development, the factors determining success have varied with each infrastructure upgrade. The difference is that a decade ago, software entered enterprises; today, it's Agents.

In the PC era, the competition was centered around hardware performance; in the mobile internet era, it revolved around traffic entry points. Over the past decade, the key word in China's cloud market has been 'scale'—whoever had more servers and more clients held greater sway. But today, what kind of cloud can support these Agents? Perhaps we can predict that in the 2026 showdown of China's AI cloud market, the winners will not be determined by catchy slogans, massive traffic, or high rankings on model leaderboards. Instead, the deciding factor will be who can truly integrate intelligence into the intricate fabric of every industry.

The biggest difference between Agents and mobile internet applications is their inherent reliance on industries. Agents need to enter real-world scenarios—whether it's bank risk control systems, power grid inspection processes, automotive R&D frameworks, or robot training platforms. Only by delving into these contexts can Agent intelligence be fully amplified, generating greater value.

Once Agents enter industries, the logic of competition shifts. For cloud providers, market share is no longer the sole metric; depth of penetration becomes the new moat. This is why, since this year, companies like Alibaba, Huawei, and Baidu have increasingly focused less on parameters and more on real-world applications. The industry now recognizes that the greatest challenge in the Agent era is not creating intelligence but integrating it seamlessly into industrial processes.

In some ways, this mirrors the early days of the mobile internet boom fifteen years ago.

The dividends of the mobile internet rise went to those who could integrate the internet into consumer, payment, retail, and logistics systems. The same holds true today: opportunities will belong to those who can bring intelligence into industries.

In the past, cloud providers sold computing power. In the future, they will deliver productivity. The game has essentially restarted, and the rules have changed. This time, only players who are deeply rooted in industries, prioritize implementation, and truly infuse intelligence into the industrial fabric will earn a seat at the center of the table.