Tencent’s results for the first quarter of 2026 have further sharpened the question around its artificial intelligence strategy: it can afford to keep spending, but it has yet to show when that spending will pay off.

The Chinese internet giant reported revenue of RMB 196.5 billion (USD 28.9 billion), up 9% year-on-year, but below analyst expectations. Net profit attributable to shareholders under IFRS rose 21% to RMB 58.1 billion (USD 8.5 billion), but missed consensus estimates. The results suggest Tencent’s core businesses remain strong, even as its AI push becomes more expensive to fund.

Management’s response was to argue that its older businesses are financing its newer bets. Chairman and CEO Pony Ma said Tencent’s core operations continued to grow engagement, revenue, and profit, generating the cash flow needed to support AI investment.

The numbers broadly support that argument. Gross profit rose 11% to RMB 111.3 billion (USD 16.4 billion), free cash flow increased 20% to RMB 56.7 billion (USD 8.3 billion), and Tencent ended March with RMB 533.7 billion (USD 78.4 billion) in cash and RMB 146.9 billion (USD 21.6 billion) in net cash. Capital expenditure rose 16% to RMB 31.9 billion (USD 4.7 billion).

But those same figures also show the cost of the AI transition. Non-IFRS operating profit rose 9% to RMB 75.6 billion (USD 11.1 billion). Excluding newer AI products such as Hunyuan, Yuanbao, CodeBuddy, WorkBuddy, and QClaw, it would have risen 17% to RMB 84.4 billion (USD 12.4 billion), and the margin would have widened to 43% from 39.9% a year earlier.

That gap is the story. Tencent’s mature businesses are generating stronger returns than the headline numbers suggest, while its AI products remain in investment mode.

Games remain its main source of cash. Domestic games revenue rose 6% to RMB 45.4 billion (USD 6.7 billion), even though the later timing of the 2026 Lunar New Year shifted some revenue recognition out of the quarter. Gross receipts grew at a teens percentage rate, driven by Honor of Kings, Peacekeeper Elite, Delta Force, and Valorant Mobile. International games revenue rose 13% to RMB 18.8 billion (USD 2.8 billion).

Advertising provided another source of strength. Marketing services revenue rose 20% to RMB 38.2 billion (USD 5.6 billion), up from 17% growth in the previous quarter. Tencent said the improvement was driven by upgraded AI-based ad recommendation models, closed-loop marketing in Weixin, the Chinese version of WeChat, and stronger spending from internet services, e-commerce, and gaming advertisers.

AI is therefore playing two roles at once. It is increasing costs through infrastructure, R&D, depreciation, and promotion. At the same time, Tencent said it is improving ad targeting, game content production, cloud demand, and user engagement.

The margin details reflect that split:

  • Gross margin in value-added services rose three percentage points to 63%, helped by a larger contribution from internally developed games.
  • Gross margin in fintech and business services rose two percentage points to 52%, reflecting a stronger fintech mix.
  • Marketing services gross margin, however, fell 0.5 percentage points to 55%, partly because of AI-related equipment depreciation and operating costs.
  • R&D expenses rose 19% to RMB 22.6 billion (USD 3.3 billion), primarily because of higher AI investment.

Tencent’s AI progress is not just a talking point. President Martin Lau said Hunyuan 3 had been deployed across 131 internal products, including Yuanbao, QQ, and WorkBuddy. He also said the model ranked first among all models on OpenRouter by token usage from April 28 and remained there after its free period ended on May 8.

Still, Lau called Hunyuan 3 “merely just the first step.” By Tencent’s own account, its AI effort remains closer to the start of the investment cycle than to the monetization phase. The company said WorkBuddy and CodeBuddy are seeing strong organic growth and retention, but it did not disclose standalone revenue or profit for either product.

Recent media coverage has reflected the same concern. Bloomberg said Tencent’s revenue miss increased pressure for a faster AI payoff, while Reuters noted that it is trying to catch up with rivals including ByteDance and Alibaba after launching Hunyuan 3.0.

The competitive backdrop is tightening. In March, Tencent’s plan to increase AI investment in 2026 became public after it said chip export restrictions had constrained spending plans last year. It also integrated WeChat with the OpenClaw AI agent, putting its messaging platform in more direct competition with offerings from Alibaba and Baidu in China’s agentic AI market.

Tencent’s Weixin ecosystem gives it a meaningful advantage, but not a guaranteed one. Lau said AI agents could eventually access the app’s mini programs as skills, allowing them to operate across Tencent’s app-within-an-app ecosystem. Tencent already controls an interface where more than a billion users communicate, shop, pay, and consume content. But management did not provide a firm timeline for when that level of integration might happen.

Investors pressed management on returns. Chief strategy officer James Mitchell said capital expenditure would rise substantially, especially in the second half of the year, as more China-designed ASICs (application-specific integrated circuits) become available. He said AI KPIs vary by use case: Games and advertising are judged by revenue and profit, newer AI products by capability and usage, and cloud by revenue and market share.

That distinction is useful, but it also leaves room for ambiguity. AI in ad technology can have a shorter payback period because better targeting can improve click-through rates, revenue, and profit. AI model training is a longer-cycle franchise investment. Lau said model training is “an investment for the future” and is unlikely to generate immediate returns.

Tencent is better placed than most to absorb that burden. Its free cash flow remains strong, its balance sheet is liquid, and it continued to buy back shares during the quarter. Management also said it expects to sustain buybacks through the rest of the year, supported by its cash-generating businesses and the sale of parts of its investment portfolio.

Even so, the harder questions remain: when will AI generate profit on its own, and will that happen before the cost curve rises too far?

Tencent’s legacy businesses are buying it time. Games are growing, advertising is accelerating, cloud demand is improving, and Weixin remains a distribution advantage that few rivals can match. The next phase, however, will require more than evidence that Tencent is serious about AI. It will require evidence that AI can eventually pay for itself.

Note: RMB figures are converted to USD at rates of RMB 6.81 = USD 1 based on estimates as of May 14, 2026, unless otherwise stated. USD conversions are presented for ease of reference and may not fully match prevailing exchange rates.