Tencent's traditional business keeps printing money while DeepSeek s fundraising amid ballooning data center costs
Two very different stories about AI spending in China are playing out at the same time. Tencent, the $500B-plus conglomerate behind WeChat and a gaming empire, posted solid revenue growth in its most recent quarter. DeepSeek, the AI startup that grabbed global attention with its cost-efficient models, has d follow-on fundraising as the bill for scaling its infrastructure climbs faster than expected.
Tencent’s core business is subsidizing its AI ambitions #
Tencent’s Q1 2026 results, released on May 13, showed revenues of RMB 196.5 billion, a 9% increase year-over-year. That number actually came in below analyst expectations.
The more telling figure sits in Tencent’s operating profits from its core activities, excluding new AI products. That metric rose 17% to RMB 84.4 billion. Gaming, advertising, and cloud services are not just holding steady—they’re accelerating.
Net profits landed somewhere between RMB 58.1 billion and RMB 69.8 billion. The gap between strong operating profits and comparatively compressed net profits reflects one thing: Tencent is spending aggressively on AI.
The company’s AI capital expenditure hit RMB 31.9 billion in Q1 alone, roughly $4.4 billion at current exchange rates. Management has signaled a substantial increase in AI capital spending for the second half of 2026.
Tencent’s Q2 2026 earnings, due August 12, will be closely watched. Analysts are projecting approximately 10% year-over-year revenue growth.
DeepSeek’s growth outpaces its wallet #
DeepSeek secured approximately $7.4 billion in external funding, pushing its valuation beyond $50 billion as of June 2026. The company has since d follow-on fundraising because data center infrastructure costs are eating through cash faster than originally projected.
The challenge is compounded by US restrictions on chip exports to China, which limit access to the most advanced semiconductors from Nvidia and other American chipmakers, forcing Chinese AI labs to either stockpile older hardware, develop domestic alternatives, or find creative architectural workarounds.
DeepSeek’s founder, Liang Wenfeng, has imposed a notable condition on investors: no poaching talent from the company.
The widening divide in China’s AI landscape #
Tencent and DeepSeek represent two poles of the same industry. One has decades of cash-generating businesses and can treat AI spending as a line item within a diversified operation. The other has a brilliant technical team, a $50 billion valuation, and a very real question about how long the money lasts.
What makes the Chinese market distinct is the chip export restriction layer. US companies can buy whatever Nvidia ships. Chinese companies cannot. That constraint raises the capital intensity for every Chinese AI lab, because workarounds require their own investments on top of baseline infrastructure costs.
For Tencent, the 17% growth in core operating profits gives management significant cover to keep spending. For DeepSeek, pausing fundraising invites questions about burn rate and runway in a market where Tencent, Alibaba, Baidu, and ByteDance are all pouring tens of billions into AI. Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our