Alibaba, Baidu plan to raise AI cloud prices up to 34%

Alibaba Cloud and Baidu Cloud said they would raise prices for some AI cloud services up to 34% from April 18 as AI demand and infrastructure costs rose.

Alibaba said services on its AI chips would rise by 5–34% and its Cloud Parallel File Storage would jump by 30%.

Baidu said its AI computing services would rise by about 5–30% and its parallel file storage would increase by 30%.

Cui Tingting, IDC China research manager, called the moves a reasonable response and said demand for computing power had surged in the global AI race.

The US cloud firms raised prices earlier this year, with Google increasing some fees by up to 100% from May 1 and Amazon Web Services raising some EC2 capacity limits for ML by about 15%.

Omdia reported Alibaba Cloud led China’s AI cloud market in H1 2025 with a 36% share.

🔗 Source: South China Morning Post

🧠 Food for thought

Implications, context, and why it matters.

China’s cloud giants move from a long-running price war toward price hikes

For years, China’s cloud market ran on aggressive discounting in a “price-for-volume” competition to gain market share 1.
Analysts at Morgan Stanley say the coordinated price increase breaks a “deflationary curse that has lasted for twenty years” 2.
Tencent raised prices on its Hunyuan foundation models (large AI models trained on broad data that can be adapted to many tasks) on its agent developer platform (a service for building and deploying AI agents) by more than 400% 3.
Alibaba Cloud lifted prices on several computing card products powered by its in-house Zhenwu 810E chip, plus a 30% increase for its Cloud Parallel File Storage service 1.

The focus shifts from selling server time to selling AI-powered agents

Rising “Token usage” (how much text an AI model processes) tracked with Alibaba Cloud’s model-as-a-service business, Bailian (Alibaba Cloud’s platform for accessing AI models via APIs), which logged its highest historical growth rate from January to March this year 4.
Analysts describe a move away from renting raw computing power toward “workstations for digital Agent employees” 4.
Cloud providers are rushing to support new agent frameworks (software tools for building AI agents) to capture the entry points and distribution for the next wave of AI applications 4.
Limited supply of advanced chips has created a “seller’s market,” giving cloud companies more pricing power 4.