Alibaba raises prices for AI compute and data storage by up to 34% amid surging demand

Alibaba has raised prices for its AI compute and data storage products, with increases in some categories reaching 34%. The company attributes the move to a sharp rise in demand and more expensive infrastructure. For the market, it is one more sign that capacity for generative AI remains scarce and keeps getting pricier.

Почему выросли тарифы

The price hike looks like a response to two trends at once. On one side, demand for compute to train and run AI models keeps growing faster than providers can add capacity. On the other, the infrastructure itself is getting more expensive: servers, accelerators, networking gear, power, and cooling all cost more, especially for clusters built for heavy AI workloads. In effect, Alibaba is passing part of these costs on to customers who already use its cloud and related services.

For large players, this comes as no surprise. Over recent months, the market has gotten used to the idea that AI has long been about more than models and interfaces — it also depends on a capital-intensive foundation underneath them. If demand for inference and training keeps rising, cloud providers gain room to revisit their pricing. In that sense, Alibaba's move matters not only on its own terms but as a market indicator: the AI infrastructure layer remains tightly constrained.

Что именно дорожает

The changes affect Alibaba's AI compute and data storage products. The company isn't just adjusting a single rate; it is signaling that the total cost of owning an AI stack is rising noticeably for businesses, even at the level of basic infrastructure. For customers, that means revisiting budgets and making tougher trade-offs between scaling speed and cost control. Put simply, the very foundation companies use to build internal AI services, chatbots, search, and analytics tools is getting more expensive.

• AI compute for running and serving models
• Data storage tied to AI workloads
• Overall price increase of up to 34% for some items
• Drivers: a spike in demand and rising infrastructure costs

This hits startups and teams with experimental products especially hard. When every extra model request, new data-processing pipeline, or expanded context translates directly into a bigger infrastructure bill, even a 10–20% increase quickly changes a product's economics. If the increase reaches 34%, some companies will look for compromises: optimizing inference, moving less critical tasks to cheaper capacity, or cutting storage volumes. The closer a product is to a real-time use case, the more any price increase hurts.

Что меняется на рынке

Alibaba's decision matters beyond its own customer base. It shows that cloud providers are starting to monetize the AI hype more confidently, and that the market is gradually moving past the phase when infrastructure could be scaled up with little regard for the final cost. For enterprise customers, this means AI strategy discussions will increasingly start not with choosing a model, but with how much it costs to keep that model running in production every day. The focus shifts from "ship the AI feature faster" to "can we keep it profitable after launch."

This will be especially visible in Asia, where Alibaba remains one of the key cloud infrastructure providers for local businesses. When a major player raises prices amid overheated demand, competitors get either an excuse to do the same or a chance to win over some customers with gentler pricing. Either way, the winner isn't the buyer but whoever already has spare capacity and access to capital.

Что это значит

The Alibaba story is a reminder of something simple: in the AI boom, the main constraint isn't ideas but compute. As long as demand for it grows faster than supply, infrastructure will keep getting more expensive, and businesses will have to calculate the payback of AI products far more carefully than they did a year ago.