Home / Blog / Google Q2 Earnings
ENGINEERING BLOG · 2026.07.23

Google Q2 2026 Earnings:
Cloud Surges 82%, but $205B AI Spending Spooks Wall Street

On July 22, 2026, Alphabet (NASDAQ: GOOGL, GOOG) posted a strong Q2 2026 earnings beat — then watched its stock drop nearly 4% in after-hours trading anyway. Revenue hit $119.8 billion, up 24% year-over-year. Google Cloud grew 82% to $24.8 billion. But investors fixated on one number: $195–205 billion in expected 2026 capital expenditure (capex) on AI infrastructure. This article is for tech investors, AI industry observers, cloud architects, and engineers tracking hyperscaler spend — covering financials, Cloud growth drivers, Search/YouTube AI metrics, the capex guide, Gemini 3.5 Pro delay, and the Frozen v2 chip, plus a Q3 watchlist and FAQ.

01

Heading into the print, the market had already priced a paradox: Alphabet would likely beat on revenue and EPS, but any capex raise would trigger selling. Q2 confirmed that script — and it is exactly why searches like "why did Google stock drop after beating earnings" spiked this week.

  • Growth is priced in; the bill is not: Cloud +82% and Search +17% show AI monetization is working, but the third capex guide hike to $195–205B means the market is pricing cost before incremental AI revenue.
  • Free cash flow turned negative: Q2 FCF was -$5.9 billion vs. +$5.3B a year ago. Alphabet is outspending operating cash generation to build AI data centers.
  • Flagship model overhang: Gemini 3.5 Pro was expected in June 2026 and still has no public launch date, while OpenAI, Anthropic, and xAI all shipped frontier models in July.
  • 2027 warning: Management said 2027 capex will be "significantly higher" than 2026 — with no clear ROI endpoint yet.

The market's message: growth is priced in; the bill for that growth is not.

02

Alphabet Q2 2026 financial snapshot
Metric Q2 2026 Wall St. Est. Q2 2025 YoY
Total Revenue$119.8B$116.9B$96.4B+24% Beat
Diluted EPS$9.11~$2.88*$2.31+294% Beat
Google Cloud Revenue$24.8B$22.4B$13.6B+82% Beat
Google Search & Other$63.3B$63.4B$54.2B+17% In-line
YouTube Ads$11.1B$9.8B+13%
Operating Income$40.8B$31.3B+30%
Cloud Operating Margin35.6%20.7%+14.9 pp
Q2 Capex$44.9B$22.5B+100%
Free Cash Flow-$5.9B+$5.3BTurned negative
Cloud Backlog$514B~$464B+$50B sequential

*GAAP EPS of $9.11 includes a ~$99B unrealized gain on equity holdings (Anthropic, SpaceX stakes). Adjusted EPS excluding this one-time item was approximately $2.88 — still in line with estimates.

Q1 2026 vs Q2 2026 YoY growth
Segment Q1 2026 YoY Q2 2026 YoY Trend
Total Revenue+22%+24%Accelerating
Google Cloud+63%+82%Sharply accelerating
Google Search+19%+17%Slight deceleration
YouTube Ads+11%+13%Reaccelerating
Operating Margin36%34%Capex/R&D pressure

03

Google Cloud was historically good this quarter: revenue accelerated from +48% in Q4 2025 and +63% in Q1 2026 to +82%, with operating income of $8.8 billion (vs. $2.8B a year ago) and cloud operating margin at 35.6%. The $514 billion backlog (up $50B quarter-over-quarter) gives unusually strong revenue visibility — more than 5 years of current quarterly run-rate.

What's driving Cloud growth:

  • Enterprise AI infrastructure — TPU and GPU compute rentals
  • Gemini Enterprise — nearly 90% of the Fortune 100 now uses it
  • Core GCP services — traditional cloud workloads still growing
  • First TPU system sales revenue recognized from hardware delivered to customer data centers
  • Wiz acquisition — cloud security integration boosting the pipeline

CEO Sundar Pichai on the call: "Q2 was an amazing quarter, with Alphabet revenues growing 24% year-over-year and Google Cloud revenues accelerating to 82% growth, driven by demand for AI infrastructure and AI solutions."

A common fear heading into earnings: AI Overviews and AI Mode would cannibalize search ad revenue. Q2 suggests the opposite — at least for now:

  • Google Search & other: $63.3B (+17% YoY)
  • YouTube ads: $11.1B (+13% YoY)
  • Google advertising total: $81.6B
  • AI Mode: 1 billion+ monthly active users since going global in October 2025
  • Gemini App: 950 million MAU
  • Gemini API throughput: 22 billion tokens/minute (up from 16B in Q1)
  • AI features in Search sending "billions of clicks to websites every week"
  • Antigravity AI coding tool: 2.4 million weekly active users

Search query volume is growing, not shrinking. AI is expanding the search experience rather than replacing it — at least in Q2.

04

Alphabet shares dropped ~4% in after-hours trading to around $327, despite beating on both revenue and EPS. Three reasons:

1. Capex guidance got raised — again

2026 capital expenditure guidance timeline
Timeline 2026 Capex Range
Early 2026$175B – $185B
April 2026 (Q1 call)$180B – $190B
July 2026 (Q2 call)$195B – $205B

Q2 capex alone was $44.9 billion — double last year's $22.5B. Management said ~60% goes to servers, ~40% to data centers and networking. CFO Anat Ashkenazi defended the spending: demand signals today are stronger than a year ago, and Alphabet uses a disciplined ROI framework.

2. Free cash flow went negative: FCF was -$5.9 billion in Q2, vs. +$5.3B a year ago.

3. The Gemini 3.5 Pro overhang (detailed below).

Where Alphabet is spending on AI infrastructure
Investment Area Evidence
AI data centers40% of capex → facilities & networking
AI compute (servers/TPUs)60% of capex → hardware
Equity raise (June 2026)$49.6B stock + preferred issuance
Debt raise (Q2 2026)$20.3B senior unsecured notes
Cash position$242.5B in cash & marketable securities
Third-party capacity bridgeUsing external compute in Q3 — may pressure Cloud margins slightly

05

Gemini 3.5 Pro: delayed, but not dead

Google's next frontier model — Gemini 3.5 Pro — was originally expected in June 2026. It hasn't shipped. Bloomberg (July 16) reported the model is months behind schedule because coding performance fell short of internal goals. Google updated training data in late June to improve coding — results were disappointing. The model exists and works — Google confirmed it's in testing with partners alongside an upgraded Flash model — but no public launch date has been given. Raymond James analyst Josh Beck said the delay has "shifted perception from leading edge to trailing edge."

Competitive context: in July 2026, OpenAI's GPT-5.6, Anthropic's Claude Fable 5, and xAI's Grok 4.5 all launched. Google has already started Gemini 4 development and plans to ship new model versions almost monthly. Gemini 3.5 Flash Cyber — a security-focused model — did launch.

Frozen v2: Google's bet on model-specific AI chips

Frozen v2 AI server chip (The Information report)
Attribute Detail
Design approachEmbeds parts of Gemini's architecture directly into silicon
Efficiency gainEstimated 6–10x more tokens per watt vs. current TPUs
ScopeGemini-only (unlike general-purpose TPUs)
ProductionSmaller initial run; complementary to TPUs, not a replacement
Revenue timingMost TPU system sales revenue expected in 2027, not 2026

Power efficiency equals cost efficiency at Google's scale. If Frozen v2 delivers, Google can serve Gemini inference cheaper than competitors renting Nvidia GPUs — directly impacting Cloud margins long-term.

6-step framework for reading hyperscaler earnings

  1. Start with the beat/miss matrix: Compare revenue, EPS, Cloud segment, and consensus — separate GAAP from adjusted EPS (this quarter's GAAP includes a ~$99B equity gain).
  2. Track capex guide revisions: Record the early-2026 → Q1 → Q2 hike and compute capex/revenue ratio trends.
  3. Check FCF and cash reserves: Negative FCF may be cyclical during raise cycles; $242.5B cash provides a buffer.
  4. Read Cloud backlog and margin: $514B backlog and 35.6% Cloud margin are the strongest demand-side signals.
  5. Compare AI product KPIs: Gemini MAU, API throughput, Antigravity WAU vs. competitor launches like DeepSeek V4 GA and Kimi K3.
  6. Map next-quarter catalysts: Model launch dates, 2027 capex guidance, Search comps, TPU external sales ramp.

06

Q3 2026 watchlist:

  1. Gemini 3.5 Pro launch — any date announcement could move the stock
  2. Search revenue comps — CFO warned Q3 faces tougher year-ago comparisons
  3. Cloud margin impact from third-party capacity bridge
  4. Capex trajectory into 2027 — management already flagged "significantly higher"
  5. Competitive pressure — Kimi K3, GPT-5.6, Claude Fable 5 all shipping in July
  6. Frozen v2 progress — any production timeline would be material
  7. TPU external sales ramp — revenue recognition expected to accelerate in 2027

Frequently asked questions

What were Google's Q2 2026 earnings results?

Alphabet reported Q2 2026 revenue of $119.8 billion (up 24% YoY), EPS of $9.11, and Google Cloud revenue of $24.8 billion (up 82% YoY). Results beat Wall Street estimates of $116.9B revenue.

Why did GOOGL stock fall after beating earnings?

Despite beating revenue and EPS estimates, Alphabet shares fell ~4% in after-hours trading because the company raised its 2026 capex guidance to $195–205 billion. Q2 capex hit a record $44.9 billion, and free cash flow turned negative at -$5.9 billion.

Is Gemini 3.5 Pro delayed?

Yes. Google confirmed Gemini 3.5 Pro is in testing with partners but has not set a public launch date. Bloomberg reported the model is months behind schedule due to coding performance falling short of internal goals.

What is Google's Frozen v2 chip?

Frozen v2 is a Google-designed AI server chip that embeds parts of Gemini's architecture directly into silicon. The Information reported it could be 6–10x more power-efficient than current TPUs for running Gemini models.

  • Google Cloud Q2 2026 revenue: $24.8B, +82% YoY, 35.6% operating margin
  • Cloud backlog: $514B, +$50B sequential
  • 2026 capex guidance: $195–205B (third upward revision)
  • Q2 capex actual: $44.9B, +100% YoY
  • Q2 free cash flow: -$5.9B (turned negative)
  • Gemini App MAU: 950M; AI Mode MAU: 1B+

Bottom line: Alphabet's Q2 2026 earnings tell a story of AI monetization working — Cloud +82%, Search still growing, Gemini adoption at scale. But the market is pricing in the cost of staying at the AI frontier, not just the revenue it generates.

Primary sources below; capex guidance, model timelines, and financial figures may change — re-open links after publication to verify:

Official sources:

Alphabet Q2 2026 Earnings Release (PDF)

SEC Exhibit 99.1 — Q2 2026

Third-party coverage:

CNBC — Google Q2 2026 Earnings Live Updates

For teams that need consistent local Agent development without being locked to hyperscaler API quotas and latency, pure cloud API dependency still means cost volatility and vendor lock-in; shared VMs add hypervisor overhead. For production environments requiring zero-loss Apple Silicon, 24/7 uptime, stable iOS CI/CD, and AI Agent automation (running Claude Code / Codex on bare-metal Mac with multi-model API routing), ZUKCLOUD bare-metal Mac mini cloud nodes are typically the more controlled option — dedicated hardware, no virtualization tax, flexible daily/weekly/monthly billing. See pricing and order, or read the bare-metal architecture manifesto for Agent hosting logic.

Last updated: July 23, 2026