Alphabet's Q2 2026: Cloud Surges 82%, Gemini 4 Begins Training, Capex Hits $205B

Alphabet delivered a quarter that would have seemed implausible just a year ago: $119.8 billion in revenue, a cloud business growing faster than almost any major enterprise at this scale, and a net income of $112.1 billion — nearly four times what it earned in the same period of 2025. The results, reported on July 22, 2026, cemented Google Cloud’s transformation from a distant third-place cloud provider into an AI infrastructure powerhouse, while CEO Sundar Pichai used the earnings call to announce that training on Gemini 4, the company’s next flagship model, has already begun.

The Numbers

Alphabet’s Q2 2026 results beat expectations across nearly every metric:

• Total revenue: $119.8 billion, up 24% year-over-year from $96.4 billion
• Operating income: $40.77 billion
• Net income: $112.1 billion (compared to $28.2 billion in Q2 2025 — a 297% increase)
• Google Cloud revenue: $24.77 billion, up 82% from $13.62 billion a year prior
• YouTube advertising: $11.06 billion, up from $9.79 billion year-over-year
• Other Bets (Waymo, DeepMind commercial projects): $1.8 billion loss, up from a $1.24 billion loss

The Google Cloud figure is the headline number. Growing at 82% — from an already-large $13.6 billion base — Cloud is now Alphabet’s fastest-growing and strategically most important business unit. For comparison, Microsoft’s Azure grew at roughly 50% in the same period, and Amazon Web Services grew at around 38%. The gap has narrowed dramatically in 12 months.

What’s Driving Cloud

The acceleration is almost entirely attributable to enterprise AI adoption. “Q2 was an amazing quarter, with Alphabet revenues growing 24% year-over-year and Google Cloud revenues accelerating to 82% growth,” Pichai told analysts on the earnings call.

Nearly 90% of Fortune 100 companies are now using Gemini Enterprise — Alphabet’s flagship AI suite for the enterprise. The Gemini App itself has reached 950 million monthly active users globally, up from roughly 500 million a year ago. The platform processes 22 billion API tokens per minute, a figure that reflects both model usage volume and the depth of integration into enterprise workflows.

Demand has grown faster than Alphabet has been able to provision capacity. The company has been supply-constrained on AI compute for most of the past year — a problem it is now moving aggressively to solve.

Capex: $205 Billion and Rising

Alphabet raised its full-year 2026 capital expenditure guidance to a range of $195 billion to $205 billion, up from the previous guidance of $180 billion to $190 billion. The increase represents roughly $15 billion in additional infrastructure spending beyond what the company had originally planned for the year.

Management indicated that infrastructure spending is likely to continue rising in 2027 as additional data center capacity — much of it already under construction or committed — comes online. Industry analysts estimate that Big Tech firms collectively will spend more than $700 billion on AI infrastructure this year, with investment expected to exceed $1 trillion in 2027.

Investors reacted with the now-familiar ambivalence that greets Big Tech earnings: strong results alongside even stronger spending guidance. Alphabet’s stock dipped roughly 4% in after-hours trading following the report, as some investors focused on the capex overshoot rather than the revenue beat.

Gemini 4: The Next Frontier

Pichai’s most forward-looking disclosure was the confirmation that Alphabet has already begun pretraining Gemini 4, which he described as the company’s “most ambitious training effort yet.” Google has been releasing new models at a pace of nearly one per month since early 2026 — from Gemini 2.5 Flash to 3.5 Pro — and the Gemini 4 pretraining suggests the company is already building the model that will succeed its current flagship.

Google is simultaneously testing Gemini 3.5 Pro in expanded deployment. The company has not given a release timeline for Gemini 4, but given the cadence of recent releases, industry observers expect a public announcement within six to nine months.

The Gemini 4 pretraining also signals something about Alphabet’s competitive posture. OpenAI is widely believed to be developing GPT-6 or a successor to its current o-series. Anthropic released Claude Opus 5 earlier this month. Meta is pushing its Llama family toward open availability. Google, having stumbled publicly with earlier Gemini launches, appears to have found its footing — and is now racing to build the next generation while monetizing the current one.

YouTube and the World Cup Effect

YouTube’s 26% year-over-year advertising growth was boosted by a notable one-time factor: the FIFA World Cup 2026, which was hosted across the United States, Canada, and Mexico, bringing unprecedented viewership to YouTube’s streaming and highlights platform. Over 1.7 billion unique viewers worldwide watched World Cup-related content on YouTube during the tournament.

The World Cup tailwind won’t repeat in Q3, making it difficult to extrapolate YouTube’s Q2 growth trajectory. But the figure underscores YouTube’s expanding role as a live and sports video destination, a segment the company has been investing in as it competes with linear television and streaming services.

The Bigger Picture

Alphabet’s Q2 results are the latest data point in a remarkable reversal of fortune for Google’s cloud ambitions. As recently as 2023, analysts questioned whether Google Cloud could ever meaningfully close the gap with AWS and Azure. The answer, it turns out, was yes — and the catalyst was AI.

The flywheel is now apparent: Alphabet trains frontier AI models (Gemini), deploys them across consumer and enterprise products (Google Search, Workspace, Cloud), earns revenue that funds infrastructure (TPUs, data centers), and uses that infrastructure to train more capable models. At $24.8 billion in quarterly cloud revenue and 82% growth, the machine is turning faster than almost anyone predicted.

The open question is whether the capex can sustain itself. At $195–205 billion annually, Alphabet is spending at a pace that would have been unimaginable for any technology company a decade ago. The bet is that demand for AI compute — from enterprise customers building applications on Google Cloud, from developers using Gemini APIs, from governments and research institutions — will grow fast enough to justify it. Q2 2026 suggests, at least for now, that it is.