SUN, JULY 26, 2026
Independent · In‑Depth · Practitioner‑Tested
✎ Large Language Models

Google Q2 2026: Cloud Grows 82%, Gemini Hits 950M Users, CapEx Doubles to $44.9B

Alphabet Q2 2026: revenue $119.8B (+24% YoY), Google Cloud surged 82% to $24.8B — acceleration from 63% in Q1. Gemini processes 22 billion API tokens per minute. Gemini app: 950M monthly active users (50M from 1 billion). 90% of Fortune 100 use Gemini Enterprise. CapEx $44.9B (+100% YoY) — Google raised $49.6B in June for AI infrastructure. Gemini 4 already getting compute. Antigravity coding tool: 2.4M weekly users.

By AIToolsRecap July 25, 2026 6 min read 138 views
Home Articles Large Language Models Gemini Google Q2 2026: Cloud Grows 82%, 22B Tokens Per...

ALPHABET Q2 2026 — KEY NUMBERS

Total revenue: $119.8B (+24% YoY)
Google Cloud: $24.8B (+82% YoY, up from 63% in Q1)
Search revenue: $63.3B (+17% YoY)
YouTube ads: $11.06B (+13% YoY)
Operating income: +30% YoY, margin expanded to 34%
CapEx: $44.9B (+100% YoY)
Gemini API: 22 billion tokens per minute
Gemini app MAU: 950 million monthly active users
Gemini Enterprise: ~90% of Fortune 100
Antigravity (coding tool): 2.4M weekly active users
Capital raised in June: $49.6B (stock + notes) for AI infrastructure

Cloud at 82% — The AI Infrastructure Trade Is Working

Alphabet revenues grew 24% year-over-year and Google Cloud revenues accelerated to 82% growth, driven by demand for AI infrastructure and AI solutions. Nearly 90% of the Fortune 100 are using Gemini Enterprise. The 82% Cloud growth is a meaningful acceleration from the 63% growth Google Cloud posted in Q1 2026 and the 48% growth in Q4 2025. The direction is clear: enterprise AI spending is accelerating, and Google Cloud is capturing a growing share of it.

Gemini models now process 22 billion API tokens per minute and the Gemini app has 950 million monthly active users. For context: 950 million monthly active users puts Gemini 50 million away from a billion — making it likely to cross that threshold in Q3 2026. Pichai cited demand for AI infrastructure and solutions as the driver of Cloud's acceleration, noting that AI is additive to search, not a replacement, with search query growth driven by popular AI features.

The CapEx Story — $44.9B in One Quarter

Google reported $44.9 billion in capital expenditures in Q2 2026, representing 100% growth year-over-year. To put that in context: $44.9B in a single quarter equals Anthropic's entire Colossus 1 compute deal ($1.25B/month × 36 months = $45B) spent in 90 days. Alphabet raised $49.6 billion by issuing stock in June and brought in $20.3 billion from senior unsecured notes in the second quarter to support AI infrastructure expansion. Google is spending at a rate that dwarfs every other AI infrastructure investment announced in 2026 — and it is raising capital specifically to do more of it.

Alphabet CFO Anat Ashkenazi noted they are in a supply-constrained environment with strong demand, and the company is not planning to return to equity markets except for addressing standard employee compensation. The supply constraint is the limiting factor — Google can spend more, and the demand is there. The bottleneck is building data center capacity fast enough to serve it.

Gemini 4 — "Very Ambitious Effort", Already Getting Compute

Google CEO Sundar Pichai confirmed the company is already dedicating resources to its Gemini 4 model as AI competition heats up. "It's a very ambitious effort," he said. The frontier is dynamic, and we are committed to being at the forefront. We are iterating on models like the 3.6 Flash, which has shown significant improvements. We are confident in our ability to compete at the frontier level.

Pichai's comments on Gemini 4 are notable given Gemini 3.5 Pro's repeated delays. Rather than defending the delays, Google is positioning Gemini 4 as the strategic leap — implying the company may be skipping iterative 3.5 Pro improvements in favour of the next generational upgrade. For enterprise buyers currently evaluating Gemini 3.5 Pro vs Claude Fable 5 or GPT-5.6 Sol: the 3.5 Pro delay may never resolve in favour of a standalone release, with its capabilities instead absorbed into Gemini 4.

Antigravity — Google's Coding Tool at 2.4M Weekly Users

Developers are increasingly adopting Google's AI coding tool Antigravity, which now has over 2.4 million weekly active users. Pichai noted that in Antigravity, users can select models from Google and from Anthropic, along with an open model OpenAI released last year. Options from Chinese labs DeepSeek, Moonshot AI and Z.ai are not available. The Chinese lab exclusion is significant given that Kimi K3 is now the strongest open-weight coding model available. Antigravity offering Claude (Anthropic) models alongside Gemini models reinforces the Anthropic-Google partnership — and makes Google's $44.9B CapEx quarter relevant to Anthropic's model training capacity.

What the Numbers Mean for the AI Market

AI is already generating revenue at scale. 82% Cloud growth driven explicitly by AI demand is not a forecast — it is a reported result. Enterprise AI spending is real, accelerating, and concentrated in the providers with established cloud relationships and compliance infrastructure.

CapEx at $44.9B/quarter is setting a new baseline. At this spending rate, Google is building data center capacity at a pace that will compound over 2-3 years. The supply-constrained environment means demand is real — the infrastructure Google is building now will become the revenue base in 2027-2028.

Gemini at 950M MAU is a distribution moat. OpenAI has ChatGPT, Anthropic has Claude. Google has Gemini at 950M monthly active users — a user base that dwarfs any other AI product. Converting even a small percentage of those users to paid Google AI Pro ($19.99/month) or Ultra ($99.99/month) is a significant revenue opportunity that does not depend on winning the model quality race.

Sources: Alphabet Q2 2026 Earnings Release (SEC) · CNBC · 9to5Google · Yahoo Finance · Benzinga · TradingKey · Related: Gemini 3.6 Flash full review → · Gemini 3.6 Flash vs Claude Sonnet 5 →

Tags
GeminiAI NewsGenerative AIGoogle2026

Spot an inaccuracy?

We verify facts before publishing and correct errors promptly. If something in this article is wrong or outdated, let us know.

Report an error →