In-depth interpretation of the AI ​​financial reports of the five giants in Silicon Valley: Google’s profit margin doubled, AWS increased its speed by 28%, and AI entered the harvest period

Core conclusion

In Q1 of 2026, the five major technology giants Google, Microsoft, Amazon, Meta, and Apple have successively handed over the first "report card" in the AI era. The core signal is clear: AI investment has moved from "storytelling" to "settling accounts".

• Google Cloud profit margin doubled: operating profit margin soared from 17.8% to 32.9%, self-developed TPU is the biggest contributor
• **AWS reverses doubts about "falling behind": 28% year-on-year growth, AI annualized revenue exceeds US$15 billion
• Microsoft AI annualized revenue is US$37 billion: Copilot has 20 million paid seats, and it has become a fact that enterprises pay the bill
• Meta expansion caused market concerns: CAPEX was raised to 125-145 billion U.S. dollars, falling by more than 10% after hours.
• The total AI capital expenditures of the five giants in 2026 will be approximately US$720 billion, a year-on-year increase of 67%

Key Points

• Event time: end of April to early May 2026 (Q1 financial reporting season) -Affected objects: AI content creators, automation practitioners, SaaS developers
• Core changes: The market shifts from "looking at the future" to "looking at the present", and AI commercialization must come up with numbers

Background: AI investment urgently needs to be realized

In the past two years, Silicon Valley giants have been "burning money" crazily on AI infrastructure - building data centers, purchasing GPUs, and training large models. Q1 of 2026 is the first window period to see returns. The inflection point of Google Cloud's profit margin curve proves that AI investment does not necessarily drag down profits; and the sharp drop in Meta's stock price after hours shows that the market no longer pays for "larger investment in the future."

Key Impact

Adaptation suggestions

1. Pay attention to Google Cloud price reduction window: After the self-developed TPU route is completed, Google has the motivation to seize the market share of AWS and Azure through price reduction. If you use GCP, pay close attention to Q2-Q3 pricing adjustments.

2. Betting on the AI ​​SaaS track: Data from companies adopting products such as Copilot and Bedrock show that the willingness to pay on the B-side is much stronger than on the C-side. Consider packaging your AI workflow as a SaaS product.

3. Optimize content monetization strategy: Meta’s CPM increased by 12%, and Google search advertising increased by 19%, which shows that AI not only improves the accuracy of advertising, but also increases the value of traffic. As we continue to deepen our SEO+ content overseas, the space for advertising monetization is expanding.

4. Be wary of "single supplier" risks: The five giants have each built an AI ecosystem, but relying on a single platform may put them at a disadvantage in future price negotiations. Maintain the flexibility of the tool chain and use automated platforms such as n8n for centered scheduling.

Specific action list

• Evaluate the currently used cloud AI services and compare the pricing changes of Google/Microsoft/Amazon
• Investigate the payment models of enterprise AI tools such as Copilot/MCP/Bedrock to determine whether your own products can be used as reference
• Check whether there is a vendor lock-in problem in the content distribution pipeline, and introduce n8n as the orchestration layer

Example: Evaluating AI API cost changes

# AI

# 100token（）

echo "=== DeepSeek V4 ==="
echo ": $0.14/token"
echo ": $0.42/token"

echo "=== GPT-4o（）==="
echo ": $2.50/token"
echo ": $10.00/token"

echo "=== Gemini（TPU）==="
echo ": $1.25/token"
echo ": $5.00/token"

Q11099，625，81%。17.8%32.9%，。TPUAI，GPU。

GeminiQ1800%，AI。19%，2400——AI，。

AWS

AWS Q1376，28%，15。BedrockQ1token，170%。AI150，Trainium200，2250。

AI370

Azure40%，AI370，123%。CopilotQ12000，74。CFO，AI。

Meta

Meta Q1563，33%。CAPEX1150-13501250-1450，10%。，——。

Q21112，Services310。AI，Apple Intelligence。AI，。

3800AI、2026AI、28%。。，（/），CAI——。

， tools OpenAI、ChatGPT、DeepSeek、Gemini、Claude、n8n、Copilot、Bedrock

Reference sources

• 36Kr authorized reprint: Emphasis on Next "The "gold content" of AI in Silicon Valley manufacturers begins to be graded"
• The Verge: Google/Microsoft/Amazon Q1 2026 Earnings Reports
• Minutes of each company’s Q1 2026 earnings conference call

Related reading

• Want to save AI fees? Watch: DeepSeek V4 vs Claude Code: 90% Cheaper, Same Quality
• Getting Started with AI Automation for Beginners: How to build an AI content automated distribution system with n8n + ChatGPT: a complete 30-minute tutorial
• Real AI money-making case: A real case of a data analyst using Claude Code + n8n to build an automated report SaaS with a monthly income of $3,800