Google Cloud Revenue Jumps 82% as AI Spending Hits $205 Billion

Alphabet reported Q2 2026 revenue of $119.8 billion on July 23, comfortably ahead of Wall Street's $116.9 billion forecast. Google Cloud was the standout, growing 82% year-on-year to $24.8 billion. But the headline numbers were quickly overshadowed by two things: a raised AI spending plan of up to $205 billion, and the company's first-ever negative free cash flow quarter, at negative $5.9 billion. Alphabet shares fell roughly 3% in after-hours trading.

What happened

Metric	Result
Total Q2 revenue	$119.8 billion (est. $116.9B)
Advertising revenue	$81.6 billion (est. $81.1B)
Adjusted EPS	$2.85 (est. $2.89)
Google Cloud revenue	$24.8 billion, up 82% YoY (est. ~64% growth)
Free cash flow	Negative $5.9 billion
AI capex guidance (revised)	$195 billion to $205 billion
Previous capex guidance	$180 billion to $190 billion
Stock move (after-hours)	Down approximately 3%

Alphabet’s April-to-June quarter was, by most measures, a strong one. Revenue cleared estimates by nearly $3 billion. Advertising, still the company’s largest business, came in ahead of expectations. Google Cloud posted what the company called its strongest growth ever, driven by businesses building and deploying AI applications at scale.

Google also recognised revenue from direct sales of its Tensor Processing Units (TPUs, the company’s in-house AI chips built to compete with Nvidia’s GPUs) for the first time. CFO Anat Ashkenazi noted that most revenue tied to those chip agreements is expected to land in 2027.

Why is Google’s stock down despite beating estimates?

The market’s reaction came down to two disclosures. First, Alphabet raised its capital expenditure guidance from $180-$190 billion to $195-$205 billion for the year. Ashkenazi also signalled another significant capex increase in 2027. Second, the company reported negative free cash flow of $5.9 billion, the first time in its history it has burned cash in a quarter.

Thomas Monteiro, senior analyst at Investing.com, put it plainly: “As long as revenue keeps accelerating, investors will tolerate it. But capital has a real cost again, and the room for error is shrinking every quarter.”

Alphabet is not alone. Big Tech firms collectively are expected to spend well over $700 billion on AI infrastructure this year, and every earnings call is now partly a negotiation between management and investors over how long the spending will be tolerated before it converts to profit.

The Gemini delay adds another question mark

Alongside the spending concerns, analysts pressed CEO Sundar Pichai on the delayed launch of Gemini 3.5 Pro, the flagship model Google had positioned as a step forward in AI coding and autonomous agents. The delay feeds a narrative that Google is ceding ground to OpenAI, Anthropic, and fast-moving Chinese AI companies, all of which have been shipping enterprise-focused models quickly.

Pichai acknowledged the gap. “There are areas where we’ve acknowledged we need to improve; coding and agentic coding is an example of that,” he said. He confirmed that testing of Gemini 3.5 Pro continues, that training on Gemini 4 has already started, and that significant compute is being directed toward the next generation.

“We are both very committed and very confident of being at the frontier for the next generation,” Pichai added. Whether that confidence is justified is the question the market is pricing in right now. For context on what AI agent benchmarking actually looks like in practice, our EdgeBench analysis covers how these capabilities are evaluated across tasks.

Our take

The 82% cloud growth number is genuinely striking. For comparison, analysts expected around 64% growth. That gap suggests Google’s AI infrastructure build-out is pulling in real enterprise spend, not just pilot projects. The TPU revenue recognition, even if small now, is also worth watching: it means Google is starting to compete with Nvidia in the hardware market, not just use Nvidia’s chips.

The cash flow situation is the real story, though. A company that has printed cash for two decades just had a negative free cash flow quarter. That is not a crisis, but it is a signal that the infrastructure race is expensive enough to stress even Alphabet’s balance sheet. Ashkenazi’s comment that demand still outpaces capacity, even after significant investment, tells you this spend is not slowing down soon.

For businesses using Google Cloud or building on Google’s AI tools, the practical implication is that supply constraints are real. If you are waiting on quota increases or trying to scale AI workloads, expect that pressure to continue into 2027. If you are evaluating AI integration options for your business, it is worth stress-testing your vendor’s capacity, not just their benchmark scores.

The Gemini delay is a genuine competitive risk, not just a PR problem. Coding and agentic tasks are exactly where enterprise buyers are placing bets right now. Every quarter Gemini 3.5 Pro ships late is a quarter OpenAI and Anthropic spend deepening developer loyalty.

What to do about it

1. Audit your current cloud AI spend now, before 2027 price adjustments from any provider hit your budget.
2. If you rely on Google Cloud AI APIs, check your quota headroom and request increases proactively rather than waiting for a capacity crunch.
3. Keep a shortlist of alternative providers (AWS Bedrock, Azure AI, Anthropic’s API) so you can shift workloads if costs spike or capacity tightens.
4. Track Gemini 3.5 Pro’s release date. If your use case centres on coding or AI agents, consider running a parallel evaluation with a competing model now rather than waiting.

The bottom line: Google’s AI business is growing fast, but so is the bill. Budget accordingly.

Source: Bing News · Google Gemini