The September 22 snapshot is striking. Market data compiled by CompaniesMarketCap puts Nvidia at roughly $5.526 trillion, Alphabet around $4.248 trillion, Microsoft at $3.697 trillion and Amazon at approximately $2.750 trillion.

Those figures value entire companies, not just their AI businesses. Google advertising, Office, Windows, e-commerce, social networks and many other businesses are all buried inside these market caps. That makes direct comparisons imperfect, but it also explains why the largest AI players can carry valuations that pure AI companies cannot easily match.

The numbers quickly become enormous

  • Nvidia: about $5.526 trillion
  • Alphabet: about $4.248 trillion
  • Microsoft: about $3.697 trillion
  • Amazon: about $2.750 trillion
  • Meta: about $1.876 trillion
  • Broadcom: about $1.728 trillion
  • AMD: about $1.018 trillion
  • Oracle: about $451 billion
  • CoreWeave: about $48 billion

These are moving market values, not permanent rankings. A single trading session can shift them by tens or even hundreds of billions of dollars.

Nvidia still stands apart. Its market value is more than five times AMD's and roughly 115 times CoreWeave's. Investors are not only paying for GPUs, but for networking, systems, software and the CUDA ecosystem surrounding them.

AMD has entered the trillion-dollar club

AMD crossed the $1 trillion market-cap threshold in September. Reuters linked the move to renewed enthusiasm for AI infrastructure and growing confidence in AMD's effort to compete across complete AI systems rather than individual chips alone.

The speed of that repricing is notable. CompaniesMarketCap put AMD at roughly $350 billion at the end of 2025. By September 2026, the figure had climbed to just over $1 trillion.

Broadcom remains larger at roughly $1.728 trillion. Its exposure comes through a different combination of networking silicon, connectivity and custom accelerators used by large cloud customers.

Cloud giants are being valued on much more than models

Alphabet and Microsoft are the public companies sitting closest to Nvidia among the largest direct AI participants. Their advantage is distribution as much as technology: cloud infrastructure, enterprise software, consumer products and enormous existing customer bases.

Amazon is worth roughly $2.75 trillion, with AWS giving it a central position in AI infrastructure. Meta stands near $1.88 trillion while spending heavily on models, data centers, custom hardware and consumer AI products.

Combined, Nvidia, Alphabet, Microsoft, Amazon, Meta, Broadcom, AMD, Oracle and CoreWeave represent more than $21 trillion in public-market value. That should not be confused with the size of the AI market itself. Most of those companies generate substantial revenue elsewhere.

OpenAI and Anthropic require a different ruler

OpenAI and Anthropic are private companies, so they do not have public market capitalizations.

Recent financing discussions nevertheless provide an intriguing comparison. Reuters has reported talks around an OpenAI valuation exceeding $1.2 trillion, while the Financial Times recently cited an Anthropic valuation of roughly $965 billion.

Those figures are private financing valuations. They are negotiated between a limited number of investors and should not be treated as equivalent to a continuously traded market capitalization.

The scale is still remarkable. Anthropic's private valuation is approaching AMD's entire public-market value and sits well above Oracle's. An OpenAI valuation above $1.2 trillion would place it in territory occupied by some of the world's largest listed technology businesses.

CoreWeave sits at the other end of the spectrum

CoreWeave is deeply embedded in the AI infrastructure story yet carries a market cap of only about $48 billion. That is tiny next to the trillion-dollar platforms supplying chips, cloud services or distribution.

Oracle, at roughly $451 billion, occupies another position in the chain. The market is clearly assigning different prices to model developers, compute providers, chip suppliers and companies that own access to end users.

That makes the AI boom less like one trade and more like several overlapping bets. Wall Street is paying an extraordinary premium for the supplier sitting at the infrastructure bottleneck, several trillion dollars for platforms capable of monetizing AI across huge customer bases, and much less for capital-intensive providers whose debt and infrastructure costs are immediately visible.

Nvidia is still the outlier

At roughly $5.526 trillion, Nvidia is worth about $1.8 trillion more than Microsoft. Its valuation is so large that ordinary comparisons quickly start sounding ridiculous.

There is a simple message buried inside that number. Investors still place the greatest premium on the company collecting money from nearly every branch of the AI arms race. The winning model may change, but much of the compute still passes through Nvidia's ecosystem.

That valuation also leaves very little room for disappointment. AI has already rewritten the stock-market hierarchy of technology companies. The profits now have to catch up with the prices.