Alphabet Q2 2026 Capex Hits $44.9 Billion as Google Cloud Growth Accelerates

Alphabet reported $44.924 billion of purchases of property and equipment in the second quarter of 2026, almost exactly double the $22.446 billion reported a year earlier. For the first six months of 2026, that figure reached $80.598 billion, versus $39.643 billion in the first half of 2025.

Those are unusually large numbers even in the current hyperscaler investment cycle. They also need careful labeling. Alphabet reports purchases of property and equipment for the whole company; the line is not a pure measure of AI chips, data-center capacity, or Google Cloud investment.

This analysis uses Alphabet’s official Q2 2026 earnings release, published July 22, 2026. The source figures are unaudited and can be supplemented by subsequent SEC filings.

The Q2 numbers

Metric	Q2 2026	Q2 2025	Year-over-year change
Purchases of property and equipment	$44.924B	$22.446B	+100.1%
Google Cloud revenue	$24.768B	$13.624B	+81.8%
Google Cloud operating income	$8.814B	$2.826B	+211.9%
Consolidated revenue	$119.796B	$96.428B	+24.2%
Net cash from operating activities	$39.069B	$27.747B	+40.8%
Free cash flow	-$5.855B	$5.301B	Not meaningful

Capital expenditure grew faster than both company revenue and operating cash flow. On Alphabet’s own non-GAAP reconciliation, Q2 free cash flow was negative $5.855 billion after subtracting purchases of property and equipment from cash generated by operations.

The trailing-twelve-month picture is less dramatic but still substantial: Alphabet reported $132.402 billion of purchases of property and equipment and $53.273 billion of free cash flow for the period ending June 2026.

Why the Cloud result matters

Alphabet said Google Cloud revenue grew 82% to $24.768 billion, led by Google Cloud Platform activity across enterprise AI solutions, enterprise AI infrastructure, and core GCP services. Cloud operating income rose to $8.814 billion from $2.826 billion.

That combination matters because it connects a large investment program to a segment with rapidly expanding reported revenue and operating profit. It does not prove the marginal return on the latest quarter’s capital expenditure. Data centers and servers have different useful lives, capacity can take time to come online, and the company does not disclose a project-level mapping from each dollar of capex to Cloud revenue.

The responsible conclusion is narrower: Alphabet is spending at a much higher rate while its Cloud segment is also growing much faster. Whether the investment earns an attractive long-run return remains an open question.

Financing adds another infrastructure signal

Alphabet disclosed that it raised $49.6 billion of net proceeds in June through a combination of Class A and Class C common stock and mandatory convertible preferred stock. The company said the proceeds were for general corporate purposes, including capital expenditures to scale AI infrastructure and global compute.

It also issued senior unsecured notes for $20.3 billion of net proceeds during the quarter. Those debt proceeds were described more broadly as being for general corporate purposes, so they should not be relabeled as a dedicated AI-infrastructure financing.

This distinction is important in our evergreen Google capex tracker and cross-company AI infrastructure capex tracker. Recognized capital expenditure, financing proceeds, annual guidance, purchase obligations, leases, and contingent commitments describe different economic facts. Adding them together would overstate the information in the disclosures.

What the capex line does not tell us

The $44.924 billion figure is compelling, but it cannot answer several questions by itself:

• How much went to servers versus data centers, networking, land, offices, or other property and equipment?
• How much capacity supports Google Cloud customers versus Search, advertising, DeepMind model development, or Other Bets?
• How much of the year-over-year increase reflects equipment prices, payment timing, or construction milestones?
• Which suppliers and geographies received the spending?

Alphabet’s investor materials say timing of cash payments can create quarterly variability. That is one reason we retain quarterly, year-to-date, trailing-twelve-month, and full-year views instead of treating one quarter as a standalone run rate.

How to follow the next layer

Company disclosures show the demand side of the infrastructure buildout. Official supplier and trade datasets provide different, imperfect views of activity farther upstream:

• The Taiwan AI supply-chain monitor tracks monthly revenue across companies with reviewed foundry, packaging, server, power, cooling, PCB, and component relationships.
• The DRAM and NAND memory-cycle monitor pairs Bank of Korea export-price indices with KITA trade values and reported weight.
• The Alphabet stock page tracks a separate question: what leading AI models say about Alphabet as an investment.

None of these datasets should be forced into one score. Supplier revenue can reflect price and mix. Trade weight is not bit volume. Company capex is not installed compute. The useful research question is where the layers confirm each other—and where they diverge.

Bottom line

Alphabet’s Q2 release establishes three facts worth monitoring:

1. Quarterly purchases of property and equipment doubled to $44.924 billion.
2. Google Cloud revenue grew 82% and Cloud operating income more than tripled.
3. Alphabet raised equity capital partly to scale AI infrastructure and global compute.

That is strong evidence that the physical AI buildout remains central to Alphabet’s strategy. It is not, on its own, evidence that every dollar is AI-specific or that the eventual return will justify the investment.

This article is for research and educational purposes only. It is not financial advice. Company disclosures can be revised or supplemented, and all calculations should be independently verified.