From infrastructure to AI: how Alibaba Cloud powers the global ambitions of Chinese companies
New tech, such as the cloud and AI, is transforming the landscape of international trade, and Chinese firms that leverage it stand to benefit greatly
New technology is transforming international trade and, with the help of the cloud and artificial intelligence (AI), Chinese firms are increasingly well placed to compete globally.
“For Chinese companies going overseas, the requirement has evolved from just providing infrastructure to infrastructure plus AI technology,” says Han Hongyuan, vice-president of Alibaba Cloud Intelligence Group. “They also want to use AI to support their business in overseas markets.
“Even before the AI wave, they were looking at how to leverage cloud providers’ big-data capabilities to help them process data and run their digital operations more efficiently,” he adds. “It’s often easier to adopt these mature technologies from cloud providers, rather than build everything themselves.”
A 2025 report by global consulting and market research firm Frost & Sullivan ranks Alibaba Cloud first globally on its growth index and second on its strategy index in serving Chinese enterprises’ overseas cloud adoption. At the same time, Alibaba Cloud’s overseas service footprint is still expanding compared with that of longer-established international providers, reflecting its role as a cloud platform built around the needs, pace and operating realities of Chinese companies going global.
Thriving on the world stage
Chinese firms expanding overseas increasingly face a common operational challenge: how to run a single, AI-ready digital system across multiple markets while meeting local compliance, latency and customer-experience requirements.
Alibaba Cloud’s answer is a full-stack AI capability built around its Qwen suite of large language models (LLM), which comprises an open-source system of various sizes and different adaptabilities. It includes a growing ecosystem of multimodalities, designed to help more enterprises realise their AI-driven innovation goals. Qwen lowers the barriers to adopting AI by allowing companies to access, modify and deploy models easily, adapting them to their needs rather than training new models from scratch.
“With our Qwen models, we give customers the freedom on how to run the model and produce business value. They can use the model service in its simplest format, or use our infrastructure to deploy the model in a more controlled manner. They can also modify the model and then run it on their chosen infrastructure,” Han says. “We give customers the choice to run the model in their preferred ways. We also develop both the model and the supporting infrastructure together to ensure customers can run it in the most optimised way on our cloud.”
The Qwen family has been downloaded more than 1 billion times globally, with over 200,000 derivative models created by developers. It supports 201 languages. Together with its global cloud infrastructure, these models form the backbone of Alibaba Cloud’s AI platform.
The AI stack is built on an architecture that is unified with Alibaba Cloud, enabling enterprises to design systems just once and deploy them seamlessly worldwide.
According to the Frost & Sullivan report, these strengths in AI capability and unified global architecture underpin Alibaba Cloud’s performance on the strategy index. They are further reinforced by a major infrastructure push by Alibaba Group, the owner of the South China Morning Post, which has announced a three-year 380 billion yuan (US$55 billion) investment to expand global capacity and AI infrastructure.
Different needs, same platform
Chinese companies expanding globally today fall broadly into two groups with very different technology needs.
Traditional industries, such as manufacturing, focus on replicating proven systems across markets so they can expand without reinventing their digital infrastructure. By contrast, a new generation of AI-native companies builds its products and business models around the technology from the start, requiring a cloud environment designed specifically for AI development and deployment.
Alibaba Cloud provides Haier, the home-appliance maker, with a flexible computing platform that integrates both public and private cloud resources, enabling the company to manage its global operations more efficiently. Electric-vehicle maker BYD, meanwhile, uses Alibaba Cloud to support the stability of key digital systems, streamline operations and reduce costs by 5 to 10 per cent.
A newer wave of AI-native firms is expanding even faster. AIGC (artificial intelligence-generated content) players such as MiniMax and PixVerse are launching products across multiple geographies while serving large global user bases.
MiniMax, in partnership with Alibaba Cloud, has developed a globally integrated data warehouse that reduces cloud computing resource usage by 75 per cent and cuts storage costs by 40 per cent while supporting the daily processing of hundreds of terabytes of data.
PixVerse, meanwhile, has optimised its inference architecture and resource scheduling with Alibaba Cloud. Its self-developed video-generation model can now deliver 1080p high-definition video in under 30 seconds and has reached more than 100 million users worldwide.
Despite having different starting points, both legacy manufacturers and AI-native companies are converging on the same requirement: a single global digital operating system with AI at its core.
Han also highlights a pattern among Chinese electric-vehicle makers: many first build their real-time, regulation-compliant TSP (telematics service provider) system and their big-data platform with Alibaba Cloud in China, then replicate the same technology stack in overseas markets. This, he says, makes global roll-out “not only a question of price, but of reusing a proven end-to-end stack”, which in turn makes it much easier for Alibaba Cloud and its clients to work together as they expand overseas.
China-rooted service model goes global
Alibaba Cloud’s infrastructure now spans 29 regions and 94 availability zones worldwide, providing a foundation for Chinese companies expanding globally. The platform has supported more than 300,000 Chinese enterprises in their global expansion and, according to the company, more than 80 per cent of Chinese companies launching overseas operations choose Alibaba Cloud as their first option.
With headquarters and core research and development teams based in China, and local teams operating in overseas markets, Alibaba Cloud can iterate products quickly and work closely with customers during development.
Its engineers can be embedded alongside client product teams during critical phases, providing hands-on technical support as companies localise products for overseas markets. Around-the-clock service coverage across multiple time zones, combined with deep familiarity with Chinese enterprises’ internal workflows, approval processes and operational rhythms, allows Alibaba Cloud to move in sync with its customers as they scale outside China.
As Chinese companies continue to expand globally, the role of cloud and AI will only grow, the company believes, adding that for providers like Alibaba Cloud, the next phase will be about not only infrastructure but also helping customers innovate faster in every market they enter.