Original | Odaily Planet Daily(@OdailyChina)
Author | Azuma(@azuma_eth)
On the morning of July 22, Beijing time, Google's parent company Alphabet announced its second-quarter 2026 earnings after the U.S. market close.
Judging purely by the performance numbers, this is an almost perfect report card. In Q2, Alphabet achieved revenue of $119.8 billion, up 24% year-over-year, with operating profit reaching $40.8 billion, an increase of 30% — the company has maintained double-digit revenue growth for the 12th consecutive quarter, demonstrating strong continued growth resilience in its core business.
- Odaily Note: As shown in the table above, many investors saw Google's explosive Q2 EPS (earnings per share) of $9.11, emphasizing that this figure is far above last year's $2.31. However, this is mainly because it includes the valuation increase of the 14% stake in Anthropic. Excluding this related gain, EPS was only $2.85, below the market expectation of $2.95.
Among them, Google Cloud, the segment most watched by the market, delivered a report card far exceeding expectations. Last quarter, Google Cloud revenue reached $24.77 billion, a year-over-year increase of 82%, making it Alphabet's fastest-growing business segment.
At the same time, Google's traditional foundation remains solid. Google Services revenue was $94.54 billion, up 15% year-over-year. Search & other revenue within that was $63.27 billion, up 17%, and YouTube ads revenue was $11.06 billion, up 13%.
From revenue and profit to progress in AI business, Google has almost delivered all the answers investors hoped to see. However, interestingly, after the earnings release, Alphabet's stock price did not rise but instead fell nearly 3% in after-hours trading.
Why isn't the market buying it? The reason is not hard to understand — compared to "how much money Google made," the question the market now values more is: "How much does Google ultimately need to pay to win the AI era?"
Capex Surges to $200 Billion, Free Cash Flow Turns Negative for the First Time
Over the past few years, thanks to the continuous profit generation from businesses like search and advertising, Google has consistently been one of the technology giants with the strongest cash flow globally. However, as AI competition intensifies, this model is changing rapidly.
To compete for leadership in AI infrastructure, Alphabet is continuously ramping up investments. In Q2, Alphabet's capital expenditures reached $44.9 billion (exceeding the market expectation of $44.2 billion), and it raised its full-year 2026 capital expenditure guidance from the previous $180 - $190 billion to $195 - $205 billion.
Looking at the expense categories, most of the funds will be used for AI infrastructure construction such as servers, data centers, and networking equipment. This means that in just this year, Google may invest close to $200 billion betting on AI.
The massive investment is already putting new pressure on Google. In Q2, due to the rapid growth in capital expenditures, Alphabet's free cash flow turned negative for the first time — the company's operating cash flow was $39.1 billion, while capital expenditures reached $44.9 billion, resulting in a final free cash flow of -$5.855 billion.
During the investor conference call following the earnings release, Alphabet CFO Anat Ashkenazi also acknowledged the cash flow issue: "Investment in AI infrastructure will continue to exert pressure on the income statement and cash flow."
Additionally, it's worth noting that to support this ongoing expansion of AI infrastructure, Google has begun large-scale borrowing. In June this year, Alphabet completed the issuance of stock and convertible preferred shares, raising approximately $49.6 billion net. Simultaneously, the company also issued $20.3 billion in senior unsecured bonds to meet capital needs.
For a company that has long relied on its cash flow advantage to win investor favor, this is undoubtedly a significant change. Of course, the market is not opposed to Google's continued investment in AI; the real question is: "How long will it take for these investments to translate into new growth curves?"
Google Cloud Performance is Bright, But Not Enough
Fortunately, the revenue performance of Google Cloud in the Q2 earnings report can somewhat alleviate some investor anxiety.
Training models and deploying AI applications require substantial computing resources, and cloud services are the key entry point connecting AI capabilities with commercial clients. In Q2, Google Cloud revenue reached $24.77 billion, a year-over-year increase of 82%, not only far exceeding prior market expectations but also marking the fastest growth rate in recent years.
More importantly, the core drivers of cloud business growth are gradually shifting from traditional cloud computing demand towards AI infrastructure and enterprise-grade AI solutions. Alphabet stated in its earnings report that this quarter's Google Cloud growth primarily came from enterprise AI infrastructure demand, Google Cloud Platform (GCP) AI solutions, and core cloud service growth.
At the same time, Google Cloud's order backlog continues to expand. As of the end of Q2, Google Cloud's remaining performance obligations (RPO) reached $514 billion, growing further from the previous quarter. More than half of this amount is expected to be recognized as revenue within the next 24 months — at least for now, Google's investment in AI infrastructure isn't just for chasing a technology race; it has already begun translating into actual business growth.
Compared to other major AI players, Google's biggest advantage has always been its more complete industry chain, covering multiple links in AI infrastructure from the Gemini model to TPU in-house chips and the Google Cloud platform.
The current high-growth trend of Google Cloud somewhat provides a phased account for the return on AI investment... But this is still far from enough, after all, while Google Cloud's annualized revenue scale has reached approximately $100 billion, Alphabet's annual capital expenditure has surged to $200 billion.
As AI infrastructure investment continues to expand, can future revenue growth maintain sufficient speed to ultimately cover this capital investment? The market still temporarily has no answer. Judging by the after-hours decline, the market likely holds a more cautious attitude towards this.
Gemini's Intelligence, Becoming the Biggest Hidden Worry?
If Google Cloud proves Google's ability to reap commercial rewards from the AI wave, then Gemini determines whether Google can maintain its lead in this long-term competition.
Over the past few years, Google has consistently emphasized its "full-stack AI" strategy. From the underlying TPU chips and data centers to the Gemini model and the Google Cloud platform, Google aims to build a complete system covering AI infrastructure and application ecosystems.
And judging from the data disclosed in the earnings report, Gemini's user base is also growing rapidly. Currently, the Gemini App has a monthly active user base of 950 million; the Gemini model API processes approximately 22 billion tokens per minute; meanwhile, nearly 90% of Fortune 100 companies are using Gemini Enterprise.
This data proves that Google hasn't missed the AI commercialization wave. However, at the same time, market concerns about Gemini still exist.
The reason is that the core of AI competition is not just user numbers and infrastructure scale; model capability also determines ecosystem appeal. Previously, the delayed release of Google's originally planned Gemini 3.5 Pro sparked market concerns about its model competitiveness, especially in high-value application scenarios like AI programming and intelligent agents (Agents). Competitors like Anthropic and OpenAI are iterating rapidly, and Gemini seems to have fallen behind noticeably.
For Google, the most critical question right now is whether Gemini can prove itself again to still be in the first tier?
In the past, Google's biggest advantage was its globally leading search portal, strong engineering capabilities, and massive data resources. However, the rules of competition in the AI era are changing. User habits may shift from traditional search to AI assistants, and developers may prioritize more powerful model ecosystems. If Gemini cannot prove itself, then even with the most comprehensive infrastructure, Google might face the risk of its application value being captured by other models.
The AI Race Shifts Focus to "Value Realization"
Looking back, Alphabet's earnings report presents a very distinct duality.
On one hand, Google is proving that AI investment isn't just a fantasy for the capital markets. The rapid growth of Google Cloud, the continued increase in AI infrastructure demand, and the expansion of Gemini's user base all indicate that AI is gradually transitioning from a technological wave to real commercial demand. But on the other hand, the market's caution is not without reason. Annual capital expenditures approaching $200 billion, free cash flow turning negative for the first time, and highly competitive model performance all mean that Google needs to prove to investors that this AI gamble will ultimately bring long-term returns exceeding the investment cost.
And the challenge Google faces is actually a problem the entire AI industry is confronting. Over the past few years, tech giants like Microsoft, Amazon, and Meta have similarly continued to expand AI infrastructure investments. Data centers, computing power, and chip procurement have long become the main battleground for giants.
In the future, the market's focus may no longer be on who invests the most, but on who can translate capital investment into commercial value faster. For Google and other giants, the AI race may have just entered a critical phase.
