Apple surpassed Nvidia as the world’s most valuable company, but only for a few hours

On July 17, for a few hours, Apple once again became the world’s most valuable company. Nvidia shares fell as much as 4.6% in the first half hour of trading and its market capitalization dropped below $4.8 trillion, while Apple touched $4.9 trillion. It was the first time since April 2025 that the Cupertino company had reclaimed the top spot.

It lasted only as long as the trading session allowed. Nvidia pared losses near the close and finished ahead by about $6 billion, according to Bloomberg: a 0.1% gap between two companies that, according to Forbes, are worth more than the GDP of Japan, the United Kingdom, or India.

Behind the crossover are two opposing business models that came to be worth almost exactly the same. Nvidia sells the tools used to build artificial intelligence. Apple sells the devices where that artificial intelligence ultimately runs. The fact that the market values them equally says a lot about which part of the chain it is willing to pay for.

Who is the world’s most valuable company

Nvidia remains the world’s most valuable company. At the close on July 20, it was worth about $4.95 trillion versus Apple’s $4.91 trillion, a difference of $32 billion. Apple briefly surpassed it on July 17, but lost the lead before the trading day ended.

The sequence was fast. Nvidia opened lower, dragged down by the rest of the semiconductor sector, and its valuation fell below $4.8 trillion. Apple, coming off highs, touched $4.9 trillion and moved ahead. By the close, Nvidia had recovered some ground and ended the session down 2.2%, just above its rival.

The move was not limited to a single day. Over five sessions, Apple cut roughly $470 billion from Nvidia’s lead: it rose 5.84% while the chipmaker lost 3.86%. For companies of this size, the $32 billion that separates them today amounts to less than one normal day of movement.

Two opposite trajectories in 2026

Apple leads the Magnificent Seven, with a gain of 15.7% in July alone and 57.8% over the past twelve months. Nvidia is up around 7% and trades about 15% below its all-time high.

The contrast is easier to understand by looking back. Nvidia had risen 239% in 2023 and 171% in 2024. It was the first company to cross $4 trillion, in July 2025, and the first to reach $5 trillion, in October of that year. Apple, which had been the first to reach $1 trillion, $2 trillion, and $3 trillion, only crossed $4 trillion in October 2025.

What changed at Apple

For much of the AI cycle, Apple was seen as the laggard: it was not training large-scale proprietary models, and its assistant had carried two years of unfulfilled promises. That view flipped in a matter of weeks.

On June 8, Apple introduced Siri AI, a complete rebuild of its assistant, supported by models developed with Google. Early impressions were mostly positive, something that had not happened with the original Apple Intelligence announcement. It became the centerpiece of a WWDC 2026 that also served as Tim Cook’s farewell.

The second boost arrived on July 15: Chinese regulators approved the rollout of Apple Intelligence in China through the agreement with Alibaba. The stock rose 4% in a single session and hit a high of $328.73. China was the most concrete obstacle to the next upgrade cycle, in a market where Huawei is already promoting its own AI features.

That was joined by the product pipeline. Nikkei Asia reported that Apple is preparing at least five new iPhone models between the second half of 2026 and the first half of 2027, including its first foldable, and that it raised the production target for that model from seven or eight million units to ten million. Morgan Stanley calculated that this roadmap opens a path toward more than 250 million iPhones shipped in fiscal 2027, and HSBC upgraded its recommendation on the stock citing the pipeline.

The reported numbers support the case. In the quarter ended in March, Apple generated $111.2 billion in revenue, up 17% from a year earlier, with iPhone growing 22% and a gross margin of 49.3%. The next report arrives on July 30.

Why the market is punishing Nvidia even though it is growing 85%

Nvidia does not have a business problem. In its latest quarter, it posted $81.6 billion in revenue, up 85% from the previous year, and analysts expect growth close to 96% for the next one. Its sovereign AI revenue — infrastructure projects driven by governments — tripled to more than $30 billion.

What has cooled is confidence in the return on all that spending. The four major cloud providers were projecting about $650 billion in combined capital expenditure for 2026, and no generative AI company has yet proven a profitable business at that scale. Doubts intensified when OpenAI and Anthropic, two of the most valuable private companies in history, filed the paperwork to go public.

At the same time, competition emerged from within its own customer base. Google makes its TPUs, OpenAI is designing its own chip with Broadcom, and Microsoft and Amazon are moving forward with custom silicon. Every workload those customers move to their own hardware is one less workload buying GPUs.

The rotation within the sector did not help either. Wall Street shifted toward the memory and infrastructure stage of the rollout: Micron crossed $1 trillion in market capitalization in May and SK Hynix listed on the Nasdaq. The Philadelphia Semiconductor Index fell nearly 19% from its highs and, even so, outperformed Nvidia.

"I don’t see a meaningful distinction if Nvidia loses the crown. It’s likely to remain a major participant in whatever comes next," summarized Benjamin Hall, vice president of research at Segal Marco Advisors, in comments to Reuters.

The market is paying more for less growth

Here is the data point that barely appears in the coverage of the crossover. Apple trades at around 36 times its expected earnings. Nvidia, around 21 times, a multiple similar to the S&P 500. The market is paying considerably more for each dollar of profit from the company growing 17% than for the one growing 85%.

It is not a pricing error: it is a preference for a type of risk.

"Apple was seen as lagging in the AI race because it was not spending on model development, but the perception has changed," Toni Meadows, chief investment officer at BRI Wealth Management, told Reuters. "Apple is less exposed to capital intensity and better positioned to monetize AI through services, ecosystem, and hardware upgrades."

The asymmetry is concrete: in its latest fiscal year, Apple invested $12.7 billion in capital expenditure against $98.8 billion in free cash flow.

It is worth not overstating the contrast. Apple spends too. It is building AI servers at a Houston plant, signed a deal worth more than $30 billion with Broadcom to manufacture chips in the United States, and another $9 billion deal with Intel. The difference is scale and purpose: Apple buys components for products it will sell, while the big cloud providers are building compute capacity they still have to learn how to monetize.

The bill is paid by the person buying the phone

The fight for the top spot is settled on a Bloomberg screen, but it has a far more concrete consequence.

The same boom that took Nvidia to $5 trillion emptied the market for memory and storage. Data centers are absorbing DRAM and NAND chip production, prices have risen, and those chips go inside every phone, every notebook, and every console. Tim Cook was explicit in June, in an interview with The Wall Street Journal: increases are inevitable. "We are doing what we can to mitigate the huge increases being passed on to us, and we have been trying to protect our customers, but the situation has become unsustainable," he said. Apple has already raised prices to compensate, a decision Reuters points to as a risk to demand.

There is a second point that explains why Apple’s model holds up without spending like the rest. Its inference infrastructure is distributed: the Neural Engine the company has included in its chips since 2017 lives in every iPhone, iPad, and Mac sold, and part of Siri AI’s processing runs there instead of on a server. Apple does not need to build compute capacity for every new user, because the user has already bought it along with the device.

The cost of that design reappears on store shelves. Running local models requires more memory, and memory is precisely what the AI race has made more expensive. That is where the dispute between Apple and Nvidia stops being a stock-market abstraction.

What will define the fight from here

There are three concrete dates. On July 30, Apple reports earnings, the most immediate catalyst to sustain or deflate the rally. On September 1, John Ternus takes over as CEO and Tim Cook becomes executive chairman, two weeks before the traditional iPhone event. And in September, the new wave of operating systems arrives, with the rollout of Siri AI staggered by region and language.

The prediction market Polymarket assigns Nvidia a 61% probability of ending December as the most valuable company, versus 23.5% for Apple. The July 17 crossover is read there as an episode, not a regime change.

For those who follow the industry, the relevant point is not the ranking but the shift in criteria: the market has stopped rewarding the size of the investment and started rewarding the ability to charge for it. For those about to buy hardware, the signal is the opposite and much more uncomfortable: as long as memory demand remains tied to data centers, device prices have little room to fall.

Apple did not win the artificial intelligence race. For now, it won the argument that AI can be monetized without financing the infrastructure that supports it. Whether that is confirmed depends on one thing, and it is not the share price: that Siri AI works as well as it was shown to.

Frequently asked questions

What is the world’s most valuable company?

Nvidia. At the close on July 20, 2026, it was worth about $4.95 trillion, compared with Apple’s $4.91 trillion. The difference is so small that it can flip in a single session: for companies of that size, it amounts to less than one normal day of movement.

Did Apple surpass Nvidia?

Yes, but for a few hours. On July 17, 2026, Apple became worth more than Nvidia during the trading session, something that had not happened since April 2025. Nvidia pared losses near the close and finished ahead, with a lead of about $6 billion.

Why are Nvidia shares falling if the company is growing 85%?

Because the market is questioning the return on AI infrastructure investment, not Nvidia’s business. The major cloud providers have committed hundreds of billions to data centers, and no generative AI company has yet proven a profitable business at that scale. On top of that is competition from chips designed by Nvidia’s own customers.

Which company was the first to be worth $5 trillion?

Nvidia, in October 2025. Before that, it had been the first to cross $4 trillion, in July 2025. Apple was the first to reach $1 trillion, $2 trillion, and $3 trillion, and has still not reached $5 trillion.