Cloud Service Providers Announce Price Increases in Succession; Computing Power Rental Stocks See Widespread Gains

As AI applications become more widespread and OpenClaw gains popularity, the computing power service market is entering a price increase cycle. Major cloud service providers such as Alibaba Cloud, Tencent Cloud, and Baidu Smart Cloud have announced successive price hikes for AI computing power and related services.

Cloud Service Providers Announce Price Increases

On March 18, Alibaba Cloud announced on its official website that due to the explosive global demand for AI and rising supply chain costs, the procurement costs for core industry hardware have significantly increased. After careful evaluation, Alibaba Cloud decided to adjust the prices for AI computing power, CPFS (Intelligent Computing Edition), and other services starting April 18. The price of computing cards like the PingTouGe ZhenWu 810E will increase by 5% to 34%, and the CPFS (Intelligent Computing Edition) storage product will rise by 30%.

On the same day, Baidu Smart Cloud issued a notice that from April 18, prices for AI computing power, storage, and other products will be adjusted, with AI computing power services increasing by approximately 5% to 30%, and parallel file storage increasing by about 30%.

Earlier this month, Tencent Cloud preemptively adjusted some model prices, raising prices for the HunYuan series models starting March 13, with some core products seeing increases of up to 400%. The collective moves by the three leading cloud providers mark the beginning of a nationwide price increase trend.

Since the beginning of this year, overseas major cloud service providers have also raised prices for core products. On January 22, Amazon announced a 15% price increase for EC2 instances used for large model training. On January 27, Google Cloud announced price hikes for data transfer services, AI, and computing infrastructure, with increases up to 100%.

Domestic Models’ Token Usage Surges

The competitiveness of domestic large models continues to improve, achieving a historic overtaking in global traffic competition.

Data from the AI model aggregation platform OpenRouter shows that from February 9 to 15, China’s large models had a weekly token usage of 4.12 trillion, surpassing the US models’ 2.94 trillion tokens for the first time. The following week, from February 16 to 22, China’s large models’ weekly usage further surged to 5.16 trillion tokens, a 127% increase over three weeks, while US models’ usage fell to 2.7 trillion tokens. This marks a substantial breakthrough in the influence of domestic large models within the global developer ecosystem.

The traffic growth of domestic models exhibits a “clustered” explosion pattern. Data from OpenRouter shows that OpenClaw has become an important traffic entry point for domestic large models. As of March 18, the total calls for Step 3.5 Flash and MiniMax M2.5 on OpenClaw in the past month ranked second and first, respectively, demonstrating strong competitiveness. These models also rank among the top in the platform’s recent application charts. According to recent performance reports, compared to December 2025, the daily token consumption of the M2 series increased by over six times.

Guojin Securities’ research report predicts that 2026 will be a pivotal year for China’s shift from “cloud training” to a dual-driven model of “training + inference,” with the computing power gap rapidly closing under the catalysis of more modalities and broader scenarios. Consumer traffic will explode in native scenes like AI dramas and programming, combined with the accumulation of B-end niche models, jointly driving a significant increase in real-time inference computing power consumption.

Performance Growth of 20 Computing Power Leasing Stocks

On March 18, the A-share computing power leasing sector surged significantly. Stocks such as Meiliyun (000815), Yunsai Zhiliang (600602), Data Port (603881), and Aoruid hit the daily limit. Guanghua Xinwang (300383), Qingyun Technology-U, and UCloud-W rose by over 10%.

As demand for computing power explodes, the profitability of leasing companies has markedly improved. According to Securities Times and Data Treasure, 32 stocks related to computing power leasing have released their 2025 performance forecasts. Among them, 20 stocks are expected to see performance growth, with 7 projected to turn losses into profits, 8 showing year-over-year profit increases, and 5 reducing losses.

Leading industry stocks such as Xinyisheng, Runze Technology, and Zhongji Xuchuang (300308) rank among the top in growth. Xinyisheng (300502) forecasted a net profit attributable to shareholders of 9.4 to 9.9 billion yuan, up 231.24% to 248.86% year-over-year. The company stated that benefiting from continuous growth in computing power investment and rapidly increasing demand for high-speed products, sales revenue and net profit are expected to significantly increase compared to the previous year.

Runze Technology’s forecast indicates a net profit attributable to shareholders of 5 to 5.3 billion yuan, up 179.28% to 196.03%, mainly due to non-recurring gains from the successful issuance of public REITs. In August 2025, Southern Runze Technology Data Center REIT was successfully listed on the Shenzhen Stock Exchange with a scale of 4.5 billion yuan, the first of its kind nationwide and the first data center REIT on the Shenzhen Stock Exchange. The company’s non-recurring net profit increased by 5.71% to 11.33%, driven by the rapid development of the AI industry and ongoing business expansion, with AIDC business achieving leapfrog growth, boosting revenue and profits simultaneously.

From a funding perspective, some leasing stocks have recently seen increased financing. Since March, eight stocks have had net financing inflows exceeding 100 million yuan, led by Tuowei Information (002261), Kehua Data, and Xinyisheng, with 580 million, 301 million, and 256 million yuan respectively.