Nvidia vs Palantir Stock: Which Is the Better Buy Now?

Nvidia and Palantir are both sold as ways to own the artificial-intelligence buildout, but investors are buying very different machines. Nvidia sells the computing platform that makes modern AI possible. Palantir sells software that helps governments and companies turn models and data into operating decisions. One collects infrastructure spending across the AI economy; the other must prove that its software can keep capturing a larger share of customer budgets.

At the September 25 close, the price of that distinction was extreme. Nvidia traded at $225.07 and 35.2 times forward earnings, while Palantir closed at $189.67 and 120.4 times forward earnings. Palantir’s multiple was 3.4 times Nvidia’s even though Nvidia’s latest reported revenue growth was higher. That valuation gap makes Nvidia the stronger risk-adjusted stock at current prices. Palantir remains the purer bet on AI software adoption, but its shareholders have far less room for an ordinary quarter.

Nvidia vs Palantir stock: the numbers side by side

• September 25 close: Nvidia $225.07; Palantir $189.67.
• September 25 session: Nvidia +0.22%; Palantir -1.52%.
• Market capitalization: Nvidia $5.43 trillion; Palantir $455.79 billion.
• Trailing P/E: Nvidia 28.5×; Palantir 162.1×.
• Forward P/E: Nvidia 35.2×; Palantir 120.4×.
• Latest reported quarterly revenue: Nvidia $96.2 billion; Palantir $1.94 billion.
• Latest reported revenue growth: Nvidia 106%; Palantir 93%.
• Latest reported gross margin: Nvidia 75.0%; Palantir 84.7%.
• One-year share-price return: Nvidia 26.61%; Palantir 6.81%.
• Consensus price target: Nvidia $327.70; Palantir $195.57.
• Implied upside to consensus: Nvidia 45.6%; Palantir 3.1%.
• Next expected earnings date: Nvidia November 17, 2026; Palantir November 2, 2026.

Prices, market values, multiples, returns, analyst targets and earnings dates come from TECHi’s quote stack, with Yahoo Finance as the winning quote provider at the September 25, 2026 regular-session close. The companies report on different fiscal calendars, so the revenue rows compare their latest reported quarters rather than the same three-month period. Analyst targets are estimates, not promises of future returns.

The table produces an unusually clean result. Palantir has the higher gross margin, which is normal for a software company that does not manufacture chips or systems. Nvidia leads on reported growth, valuation, recent share performance and consensus upside. A smaller company can grow into a premium, but Palantir is not being valued merely as a smaller Nvidia. It is being valued as if years of near-flawless software execution are already visible.

Nvidia is converting AI spending into revenue at enormous scale

Nvidia’s latest quarter weakens the common argument that its size must force growth down quickly. The company reported $96.2 billion of fiscal second-quarter revenue, up 106% from a year earlier and 18% sequentially. Data Center revenue reached $89.0 billion, up 117% year over year. Gross margin was 75%, and non-GAAP diluted earnings were $2.22 per share.

Those figures matter because Nvidia is already a $5.43 trillion company. Doubling revenue from that base requires customers to absorb tens of billions of dollars of additional hardware, networking and software in a single year. It also shows why Nvidia should not be treated as a conventional semiconductor cycle while frontier labs, cloud providers, sovereign buyers and enterprises are adding AI capacity in parallel.

The company still has a concentration problem. Data Center supplied more than nine-tenths of quarterly revenue, and Nvidia’s latest Form 10-Q identifies land, power, data-center shells and capital as necessary constraints on customer deployments. Demand can be real while projects still slip because power, financing or construction is late. TECHi has also examined Nvidia’s growing balance-sheet exposure to customer financing. Custom accelerators from hyperscalers threaten the most standardized inference workloads.

Nvidia’s defense is broader than one GPU. Its system includes accelerators, networking, CUDA software, libraries and developer tools. A customer deciding to replace a chip must also account for software migration, engineering time and system performance. That installed base gives Nvidia pricing power, but the 75% gross margin tells competitors exactly how large the prize is.

Palantir has exceptional economics and an unforgiving stock price

Palantir’s operating performance is difficult to dismiss. Its second-quarter business update showed revenue of $1.94 billion, up 93% year over year and 19% sequentially. U.S. commercial revenue grew 149% to $764 million, while U.S. government revenue increased 90% to $809 million. The company reported $1.22 billion of adjusted free cash flow, equal to a 63% margin, and a 62% adjusted operating margin.

This is the strongest part of the Palantir case. The company is no longer dependent on slow government growth or a handful of bespoke deployments. Commercial customers are adopting its Artificial Intelligence Platform, and large contracts are following: Palantir said it closed 220 deals worth at least $1 million during the quarter, including 73 worth at least $10 million.

Palantir also sits closer to the operating decision than Nvidia does. A manufacturer can use Palantir to manage production, a defense agency can use it to coordinate missions, and a hospital can use it to allocate resources. If that software becomes part of daily operations, switching costs can rise with the customer’s dependency on workflows, data models and trained staff.

The difficulty is paying 120.4 times forward earnings for that advantage. Palantir’s $455.79 billion market value equaled roughly 74 times its trailing twelve-month revenue in the September 25 market snapshot. Even a software company growing above 90% can suffer multiple compression if growth slows from extraordinary to merely excellent. TECHi’s latest Palantir forecast reaches the same pressure point from the contract side: operating momentum is strong, but expectations rise with every award. The company’s June-quarter Form 10-Q provides the GAAP basis investors should use alongside management’s adjusted metrics.

Why the valuation gap matters more than the share prices

A $189 stock is not cheaper than a $225 stock. Share price says little without the number of shares, earnings and cash flows behind it. On market value, Nvidia is almost 12 times Palantir’s size. On quarterly revenue, it is nearly 50 times larger. Yet Nvidia’s forward earnings multiple is less than one-third of Palantir’s.

That mismatch does not mean Palantir is a poor company. It means the stock carries a more demanding contract with its owners. To justify a 120.4 forward P/E, Palantir needs sustained high growth, expanding earnings, strong cash conversion and limited dilution. A slowdown in one of those variables can hurt the multiple before it meaningfully changes the underlying business.

Nvidia’s 35.2 forward P/E also embeds risk. The company must keep hyperscaler capital spending productive, execute new product transitions and defend its software moat as customers develop internal chips. But the current multiple offers a larger earnings yield and leaves more room for growth to normalize.

Consensus targets reinforce the difference, with an important caveat. Nvidia’s $327.70 average target implied 45.6% upside from the September 25 close. Palantir’s $195.57 target implied only 3.1%. Analysts can be wrong, and the very wide ranges around both names show substantial disagreement. The comparison is still useful because it shows where expectations sit: Wall Street sees far more price room in Nvidia, while Palantir already trades near the average target.

Which company has the stronger moat?

Nvidia owns the more horizontal moat. Every major AI developer needs compute, and Nvidia can sell into model training, inference, robotics, autonomous systems and scientific workloads. CUDA and its surrounding libraries reduce the amount of work developers need to do before a model runs efficiently. The moat is strongest when the workload is complex, performance-sensitive and moving quickly.

Palantir’s moat is vertical and organizational. Its software becomes valuable when it connects data, permissions, models and decisions inside a specific institution. The more deeply a deployment enters an operating process, the less useful a generic model or dashboard becomes as a substitute. Palantir can therefore build intense customer loyalty without controlling the underlying compute.

These moats can coexist. Palantir can run workloads on Nvidia infrastructure, and Nvidia benefits when enterprise AI software creates more demand for accelerated computing. The investment question is about who captures the better return from the next dollar of adoption. Nvidia currently collects a larger share of industry spending at a much lower earnings multiple. Palantir may deliver faster per-customer expansion, but investors are paying heavily in advance for it.

What could change the verdict?

Palantir becomes the stronger choice if its U.S. commercial growth remains near triple digits while adjusted margins hold above 60% and earnings compound fast enough to pull the forward multiple down without a stock-price decline. The November 2 earnings report should reveal whether the second quarter was a durable step-up or an unusually strong cluster of contract starts.

Nvidia loses its advantage if AI infrastructure orders decelerate sharply, customer financing weakens, or custom silicon takes meaningful share before new workloads expand the market. Its November 17 report will be judged on Data Center growth, gross margin and evidence that buyers are earning enough from deployed capacity to keep spending.

There is also a portfolio question. Nvidia provides broader exposure across the AI economy but carries semiconductor-cycle, supply-chain and capital-spending risk. Palantir provides more concentrated exposure to operational AI software, government programs and commercial adoption, with a much higher valuation risk. Owning both reduces the need to predict which layer captures all the economics, but it does not eliminate the fact that both stocks can fall when AI expectations reset.

The better stock now

Nvidia is the better risk-adjusted stock at the September 25 close. It delivered faster latest-quarter revenue growth, traded at 3.4 times less forward earnings valuation than Palantir, and carried substantially more upside to the average analyst target. The result is unusual: the larger company currently offers both the stronger reported growth rate and the less demanding price.

Palantir is the more aggressive choice. Its commercial acceleration, software margins and position inside customer workflows can support exceptional long-term earnings growth. Investors buying it at 120.4 times forward earnings are also accepting that a very good quarter may not be good enough. At current valuations, Nvidia needs continued strength; Palantir needs continued exceptionalism.