Skip to main content
Updated

AI Economy Market Cap Report: Top 20 Companies Added $2.15 Trillion Since 2025

Jazib Zaman
By
• Lahore, Pakistan44 min read
Reviewed by
Muhammad Zeshan Sarwar
Muhammad Zeshan Sarwar
Fact-checked by
Nouman S. Ghumman
Nouman S. Ghumman
Share
AI Economy Wealth

The public AI economy has added a net +$2.15T of market value since the end of 2025, based on a TECHi basket of 20 companies most directly tied to AI compute, cloud distribution, models, memory, networking, semicap equipment and enterprise AI software. The same basket stood at $28.087T at 2025 year-end and $30.236T on May 6, 2026, using the company-by-company market-cap history pages linked throughout this report.

That number needs one serious caveat: this is AI-economy market-cap change, not a perfect claim that every dollar was caused by AI. Apple, Microsoft, Amazon, Alphabet and Meta move for advertising, devices, cloud, rates, regulation, margins and buybacks too. But AI is now one of the central valuation arguments for the whole basket, and the market is putting a very large number on the companies that supply the compute, software, memory, networking and distribution layer.

The strongest signal is not NVIDIA alone. It is the way value spread across the stack. Alphabet, TSMC, Amazon, Micron, Broadcom, AMD, ASML, Arm and Arista all added market value in different parts of the AI chain, while Microsoft, Meta, Tesla, Salesforce, Adobe, ServiceNow, Oracle and Palantir show the other side of the story: AI exposure does not automatically protect a stock when investors question timing, margins, monetization or valuation.

Methodology: What TECHi Counted

TECHi built this report around a public-market basket of 20 companies that sit closest to the AI economy: accelerator compute, hyperscale cloud, model distribution, advanced foundry, semicap equipment, high-bandwidth memory, AI networking, enterprise AI software, operating platforms and GPU cloud.

Market-cap figures come from CompaniesMarketCap, using each company's current market-cap page and its end-of-year historical market-cap series. The report compares the 2025 year-end market cap with the current 2026 market cap shown on the company page at the time of the May 6, 2026 research pull.

This is not a factor model. It does not isolate AI from interest rates, buybacks, product cycles, advertising growth, cloud margins, currency or regulatory risk. It answers a narrower but useful question: how much public-market value has moved inside the companies most responsible for building and distributing the AI economy?

TECHi chart showing AI economy market-cap additions and drags for the top public AI companies since 2025 year-end.

The Answer: A Net +$2.15T Added

The 20-company basket increased from $28.087T at 2025 year-end to $30.236T on May 6, 2026. The net increase was +$2.15T, or 7.7%.

The positive side was larger than the headline number. Winners in the basket added +$3.34T of market value, led by Alphabet, TSMC, Amazon, Micron, Broadcom and AMD. Drags subtracted -$1.20T, led by Microsoft, Meta, Palantir, Tesla and Salesforce.

The Top 20 AI Economy Market-Cap Ledger

1. NVIDIA (NVDA)

2. Alphabet (GOOG)

3. Apple (AAPL)

4. Microsoft (MSFT)

5. Amazon (AMZN)

6. TSMC (TSM)

7. Broadcom (AVGO)

8. Meta Platforms (META)

9. Tesla (TSLA)

10. Oracle (ORCL)

11. ASML (ASML)

12. AMD (AMD)

13. Salesforce (CRM)

14. Adobe (ADBE)

15. ServiceNow (NOW)

16. Palantir (PLTR)

17. Arm Holdings (ARM)

18. Micron Technology (MU)

19. Arista Networks (ANET)

20. CoreWeave (CRWV)

What The Market Is Actually Paying For

The AI value pool is not one business model. It is a stack.

At the bottom sits the hard infrastructure layer. NVIDIA supplies accelerators, systems, software and networking. TSMC manufactures the leading-edge silicon. ASML supplies the lithography chokepoint. Micron supplies high-bandwidth memory. Arista supplies AI data-center networking. CoreWeave rents accelerator capacity as a cloud utility.

Above that sits the hyperscaler and platform layer. Amazon Bedrock, Microsoft AI, Google Gemini and Oracle AI all represent attempts to make models, data, cloud infrastructure and enterprise deployment easier to buy.

Then comes the application and workflow layer. Salesforce Agentforce, Adobe Firefly, ServiceNow AI Agents and Palantir AIP are all trying to turn AI from a demo into work that gets executed inside customer systems.

The Standout Story Is Not Just Chips

NVIDIA added +$137.00B in this window, but Alphabet added +$853.00B, TSMC added +$475.00B, Amazon added +$457.00B, and Micron added +$401.44B. That tells us investors are not only buying the GPU story; they are buying the whole AI supply chain.

The flip side matters too. Microsoft, Meta, Tesla, Palantir, Salesforce, Adobe and ServiceNow all remain AI-critical companies, yet each posted a negative market-cap change in this measurement window. That does not mean their AI strategies failed. It means AI exposure is no longer enough by itself. Investors are now asking whether AI turns into revenue, margin, defensible product advantage or real customer adoption quickly enough to justify valuation.

Why This Wealth Was Created

The first reason is scarcity. The AI economy is constrained by accelerators, advanced packaging, HBM memory, networking and data-center power. When a scarce input becomes central to the next platform shift, public markets capitalize the bottleneck.

The second reason is distribution. Alphabet, Microsoft, Amazon, Apple and Meta already own huge customer surfaces. If AI becomes a daily interface, the companies with the largest distribution layers can monetize it through cloud consumption, ads, subscriptions, devices and enterprise software.

The third reason is workflow control. Salesforce, Adobe, ServiceNow and Palantir matter because enterprises do not pay for models in the abstract; they pay when models enter sales, service, creative, document, operations, defense and compliance workflows.

The fourth reason is optionality. Tesla, Apple and Meta carry AI optionality that is harder to measure in quarterly revenue but deeply embedded in the valuation conversation: autonomy, on-device agents, consumer AI, robotics, smart glasses and new interfaces.

What Would Change The Number Next

The next leg of wealth creation depends on whether AI moves from capex promise to cash-flow proof.

For infrastructure companies, the test is supply durability: can NVIDIA, TSMC, Micron, Broadcom, ASML, AMD and Arista keep converting AI demand into high-margin revenue without a supply glut? For hyperscalers, the test is utilization: can Amazon, Alphabet, Microsoft and Oracle earn attractive returns on the AI data centers they are building? For software companies, the test is monetization: can agents and copilots produce visible seat expansion, usage pricing, retention or margin leverage?

That is why this report should not be read as a victory lap. It is a ledger. The market has already created trillions of dollars of AI-linked value. The next phase has to justify it.

FAQ

Frequently asked questions

How much market value did the TECHi AI Economy Top 20 add since 2025 year-end?

The basket increased from $28.087T at 2025 year-end to $30.236T on May 6, 2026, a net increase of +$2.15T. The figure uses CompaniesMarketCap current and historical market-cap pages for each company.

Does this mean AI caused every dollar of the market-cap change?

No. This is an AI-economy basket measurement, not a factor model. Market caps also move because of rates, margins, regulation, non-AI products, buybacks, currency and investor sentiment.

Which company added the most market value in the basket?

Alphabet added the most in this measurement window at +$853.00B, followed by TSMC, Amazon, Micron and Broadcom.

Why are non-model companies like TSMC, ASML and Micron included?

They are included because AI value depends on the physical supply chain: advanced foundry capacity, EUV lithography, high-bandwidth memory and networking. Model labs cannot scale without that infrastructure.

Why did some AI companies lose market cap?

AI exposure does not guarantee stock performance. Microsoft, Meta, Tesla, Palantir, Salesforce, Adobe, ServiceNow and Oracle were drags in this window because investors also price growth expectations, margins, valuation, execution risk and non-AI business performance.

Share

Pick your channel

Spotted an error?Report a correction →

About the Author

Jazib Zaman
Jazib ZamanScore 59

CEO at TECHi — Markets, AI & Tech Strategy

CEO of TECHi. Building the operating system for serious tech investors. Previously led engineering at scale. Focus: AI capex thesis, semiconductor supply chain, and the equity tape.

View profileSign in to follow

Comments

Sign in to join the discussion