1. Chinese cloud providers such as Alibaba and Tencent have begun raising prices for artificial intelligence (AI) computing and storage services as unprecedented demand for AI meets escalating hardware costs, representing a significant shift in an industry that has traditionally used price cuts to compete for market share. [para. 1]2. Alibaba Cloud announced that, effective April 18, prices for several AI computing products would increase by 5% to 34%, and its cloud parallel file storage service would see a 30% price hike. The company cited global AI demand, supply-chain challenges, and rising procurement prices for core hardware as the main reasons. [para. 2]3. This pricing trend follows similar moves by major Chinese competitors. Tencent Cloud ended free public testing of certain large language models and raised prices on its Hunyuan 2.0 models by over 400%. Baidu’s cloud arm increased AI computing-related service prices by about 5% to 30%, and file storage products by roughly 30%. [para. 3]4. International rivals have acted even earlier. Amazon Web Services (AWS) hiked EC2 prices used for large model training by 15% in January, and Google Cloud raised rates for data transmission, AI, and computing infrastructure—including up to a 100% increase for data transfer in North America. [para. 4]5. Alibaba attributed the increases in part to a spike in “token” usage, which measures activity in generative-AI services—a metric surging alongside AI adoption. [para. 5]6. During the Lunar New Year, the rapid spread of AI agent applications—led by OpenClaw—fueled record growth in Alibaba Cloud’s Bailian model-as-a-service business between January and March 2026. The company is now focusing scarce AI computing resources on token-based services. [para. 6]7. Analysts, like Omdia’s Raymond Zhan, identify the sharpest price jumps in advanced AI computing and high-end storage products, driven by the limited supply of powerful chips. Increases have also affected standard cloud products. [para. 7]8. Cui Tingting of IDC China explains that global AI demand is causing providers to ramp up investments in infrastructure, reducing hardware availability and lengthening delivery cycles, thus forcing price increases. These upgrades include data centers with greater power and cooling needs—a costlier proposition than traditional facilities. Cloud companies also need funds to compete in research, marketing, and infrastructure for future growth. [para. 8][para. 9][para. 10]9. Cui adds that rising tensions in the Middle East are impacting the supply of energy, chemicals, and metals essential for core hardware, potentially worsening the already tight hardware market. [para. 11]10. Memory cost escalation is a key factor driving higher AI computing expenses. Since mid-August 2025, North American AI datacenter demand pushed memory chips—including DDR4, DDR5, and NAND flash—into a historic upswing, with DDR4 16Gb prices soaring by 1,800% in 2025, DDR5 by 500%, and 512Gb NAND flash by 300%. Prices continued rising into 2026, compounding the cost pressures on cloud providers. [para. 12][para. 13]11. This environment reverses years of aggressive discounting in China’s cloud sector, moving instead toward a more mature, less-subsidized pricing model as hardware scarcity and shifting global trends take hold. [para. 14]12. Analysts expect AI cloud service prices, especially for model training and inference, to remain high. Providers are likely to introduce further price adjustments—potentially through changes to token billing. Cost relief will depend on when global hardware supplies increase, and another price rise cannot be ruled out if pressures continue. [para. 15][para. 16]13. The rapid mainstreaming of OpenClaw-style AI agents since March 2026, with Alibaba, Tencent, and ByteDance launching consumer and enterprise products, is fueling token consumption and giving providers a fresh avenue to monetize AI. Startups such as Zhipu AI, MiniMax, Moonshot AI, and StepFun are also rolling out related services. [para. 17][para. 18]14. This shift fits well with the operational models of internet-based cloud companies, offering integration into platforms like Alibaba’s e-commerce, Tencent’s WeChat, gaming, and audio-video services. The competition for leadership in token-based services is intensifying and may define the next phase of the AI cloud market. [para. 19][para. 20]15. Looking forward to 2026, analysts expect robust growth for the Chinese cloud sector, fueled by AI adoption and international expansion. However, higher infrastructure costs mean that providers are under increasing pressure to improve not just revenue, but also profitability, marking a new era of financial discipline in China’s cloud industry. [para. 21][para. 22]AI generated, for reference only