Palantir jumps more than 10%: this time, what’s truly being repriced isn’t the AI story—it’s the pace of execution
Palantir has turned in yet another hard-to-find-fault earnings report.
Revenue in the second quarter of 2026 reached US$1.935 billion, up 93% year over year and clearly above market expectations of about US$1.8 billion; adjusted earnings per share were US$0.41, also exceeding analysts’ expectations. After the earnings report was released, the stock price rose in after-hours trading by as much as about 10%–14%. The figures provided by different media outlets vary slightly, mainly due to differences in the reporting time points. The U.S. Securities and Exchange Commission
Just looking at these numbers, it’s easy to attribute the rise to an “earnings beat.”
But what truly drove the market to reprice is not just how much more Palantir earned in a single quarter. It is that Palantir simultaneously proved three things:
Growth is still accelerating, profits are not being eaten away by growth, and management believes this pace can continue for at least a few more quarters.
When these three things happen together, that’s the real reason the stock price reacted so directly this time.
I. More important than beating revenue expectations is that growth does not show a clear slowdown.
For a software company that already has nearly $200 million in quarterly revenue, 93% year-over-year growth is unusual by itself.
More importantly, this isn’t being propped up by a one-off breakout from any single business.
Palantir’s second-quarter U.S. commercial revenue reached $764 million, up 149% year over year; U.S. government revenue reached $809 million, up 90%. Together, the two segments contributed $1.573 billion, about 81% of total company revenue. U.S. Securities and Exchange Commission
This set of data shows that Palantir currently has its feet on two growth curves at the same time.
One piece of information from enterprise customers: more and more companies are no longer satisfied with buying access to a model’s usage quota. They need to truly integrate the model into supply chains, production, risk control, healthcare, and internal decision-making processes.
Another line comes from government and defense. Geopolitical conflicts, military digitization, and upgrades to intelligence systems are causing governments to keep increasing investment in real-time data processing, battlefield decision-making, and autonomous systems. As a result, Palantir’s U.S. government business grew by 90%. Reuters
One of the biggest questions the market has had about Palantir in the past is whether it is a specialized software company that relies on government contracts, or a platform-type company that can enter mainstream enterprise markets.
Now the earnings report provides at least some interim answers: the government business remains strong, but the commercial business is moving faster.
II. What Palantir sells is not a “big model,” but getting the model into real production environments.
Palantir is often placed into valuation frameworks for AI software companies, but it isn’t the same kind of business as typical big-model application companies.
Enterprises are not short of models that can write summaries, generate code, or answer questions.
The real challenge is: how to let models read internal data without overstepping permissions; how to know where a specific piece of data comes from; how to turn model recommendations into an approval, execution, and accountability workflow; and if the model makes mistakes, who should be held responsible.
These problems can’t be solved by one chat.
Palantir’s key position is closer to the control layer between enterprise data, business processes, and models. Models can be swapped, but the enterprise’s internal data structures, permissions, business objects, and decision-making processes won’t change easily.
This is also why Palantir repeatedly emphasizes “sovereign AI.” What it wants to convey is not an abstract concept, but that customers should keep control over their own data, model choices, and final decisions, instead of handing the entire business logic to some model company.
From this perspective, as big-model capabilities continue to commoditize, it doesn’t necessarily weaken Palantir.
The more models you have and the more complex the deployment environment is, the more enterprises need a platform that can manage data, models, permissions, and execution workflows.
Of course, whether this logic can hold long term ultimately depends on whether customers are willing to continue paying. The Q2 U.S. commercial contract data provides a strong signal: total U.S. commercial contract value reached $2.132 billion, up 153%; U.S. commercial remaining contract value reached $6.238 billion, up 124%. U.S. Securities and Exchange Commission
This means enterprise customers are not merely trialing the product—they are expanding the contract size.
III. The strongest part of this earnings report is actually its profit margin.
Many software companies with high-speed growth can use heavy sales spending to generate revenue growth.
This kind of growth is not scarce.
What’s truly scarce is: while revenue is nearing a doubling, profit margins are also improving noticeably.
Palantir’s Q2 GAAP operating income was $912 million, with an operating margin of 47%. Adjusted operating margin reached 62%. Operating cash flow and adjusted free cash flow both exceeded $1.2 billion, corresponding to about a 63% profit margin. U.S. Securities and Exchange Commission
Breaking down costs gives a clearer picture.
Second-quarter revenue grew 93% year over year, but sales and marketing expenses grew only about 39%, R&D expenses grew about 43%, and G&A expenses grew about 20%. Revenue growth far outpaced major operating expenses, indicating that new contracts are starting to generate clear scale effects. U.S. Securities and Exchange Commission
The ideal state for a software company is that after the product is completed, it can be deployed repeatedly, and costs do not need to increase in proportion to year-over-year revenue.
In the past, Palantir relied for a long time on front-line engineers going deep into customers’ internal systems to complete deployments. As a result, the market has always worried it looks more like a high-end consulting firm than a standard software platform.
This change in profit margin indicates that, at least in the current stage, product delivery is becoming more standardized, and expanding existing customers is cheaper than winning new customers from scratch.
This may be more important than pure revenue growth.
IV. Raising guidance is the direct fuel for the stock price rally.
Trading in the market is never just about the earnings of the just-passed quarter, but rather the profit expectations for the next several quarters.
Palantir raised its full-year 2026 revenue guidance for 2026 from about $7.65 billion previously to $8.150 billion—$8.158 billion, increasing by nearly $0.5 billion at once. Reuters
The Q3 revenue guidance is $2.160 billion—$2.164 billion, higher than the average analyst expectation of about $2.0 billion previously. Reuters
This is not a routine small upward revision.
Management is effectively telling the market: Q2 growth was not revenue recognized early, nor was it short-term volatility from a specific government contract. The company’s current view of orders and delivery timing is stronger than what it expected three months ago.
Before the earnings were released, investors may have still been debating whether Palantir can sustain high growth.
After the earnings were released, the question became how long it can sustain high growth.
In the valuation model, revenue, profit margins, and future growth rates have all been raised at the same time, so the stock price would naturally respond quickly.
V. However, growth being delivered doesn’t mean the valuation is cheap.
Palantir’s biggest problem right now is no longer whether the company is doing well—it’s how much the market is paying for that excellence.
Based on the latest quotes, Palantir’s market cap is about $323 billion and its P/E ratio is about 141. Roughly calculated using the midpoint of management’s latest full-year revenue guidance, the market cap still equals about 40 times annual revenue. U.S. Securities and Exchange Commission
This means that what is implied in the current stock price is not “Palantir will keep growing,” but:
Companies need to maintain growth far above that of ordinary software companies for many years, while sustaining very high profit margins—and they must not show clear execution missteps.
Valuations like this are not sensitive to ordinary good news; they are only sensitive to sustained performance above expectations.
When the market has priced in a full-score valuation, even a 90-point earnings report can lead to a stock decline. As long as growth drops quickly from 90% to 40%—50%, or the conversion speed of enterprise contracts slows, the valuation multiple could shrink ahead of the underlying fundamentals.
So this rally can be understood, but you can’t simply equate it with the stock price becoming “safe” again.
VI. Contract numbers look great, but you can’t directly treat them as revenue.
Palantir’s total contract value signed in the quarter reached $3.373 billion, up 49%. It includes many large contracts worth millions of dollars. The company signed 220 contracts worth at least $1 million, including 73 contracts worth more than $10 million. U.S. Securities and Exchange Commission
But here you need to stay alert.
Palantir itself also explains in its earnings materials that total contract value and remaining contract value may include contract options that customers can choose to execute in the future, and many contracts allow customers to terminate for convenience.
In other words, contract value represents potential business scale, not that these funds are already locked in—and it also cannot be equated with current revenue. U.S. Securities and Exchange Commission
What’s truly worth watching next is not only that contract totals keep hitting new highs, but whether these contracts can consistently convert into revenue and cash flow.
Having lots of orders is one thing; customers continuing to use them and expanding deployments is another.
VII. The U.S. market is an advantage—but also a concentration risk.
In this quarter, more than 80% of Palantir’s revenue came from the United States. U.S. government and U.S. commercial business each account for nearly 40% of total revenue.
In the short term, this concentration has made Palantir a direct beneficiary of the U.S. government increasing defense and AI spending.
Over the long term, it also brings policy and regional risks.
Some European governments are stepping up their focus on local data sovereignty and domestic software vendors. France’s domestic intelligence agencies plan to replace Palantir with products from the local company ChapsVision; Palantir’s policing contract in the U.K. is also facing legal disputes. Reuters
This doesn’t mean the U.S. business will suddenly slow down.
But it also reminds investors that Palantir’s advantages in the U.S. may not be replicated as-is in other countries. Different countries have different requirements for control over defense, intelligence, and critical data, and localized procurement could become a long-term headwind to international expansion.
In addition, Palantir’s cooperation with defense, immigration, and law enforcement agencies has long been controversial. No matter what position investors take on these issues, they bring real brand, regulatory, and procurement risks—not just noise in the public conversation. The Guardian
II. Equity incentive is still worth tracking
Palantir’s stock-based compensation in Q2 was about $265 million, higher than about $160 million in the same period last year, which is roughly 13.7% of revenue for the quarter. U.S. Securities and Exchange Commission
The good news is that the 47% GAAP operating profit margin the company reported already includes these costs. In other words, this profit was not packaged entirely through non-GAAP metrics after stripping out equity incentives.
Need to note that equity incentives are still a real cost.
It may not immediately burn cash, but it will affect shareholders’ long-term ownership proportion. The year-over-year increase in diluted shares is not that large, suggesting short-term dilution has not run out of control. Still, this expenditure needs to be monitored going forward. U.S. Securities and Exchange Commission
Conclusion: the company is becoming more certain, yet the stock price is becoming more of a test of judgment.
This earnings report really deserves such a clear jump.
Both revenue and profits beat expectations, commercial and government businesses are accelerating at the same time, cash flow is strong, and full-year guidance has been raised significantly. Palantir is proving that it’s not just riding an AI narrative—it has already converted AI demand into orders, revenue, and profit.
But this earnings report also did not resolve the valuation issue.
Palantir’s underlying fundamentals are becoming more certain, yet the market is giving it less and less room to make mistakes.
Next, even more worth watching than whether “AI demand exists” is whether U.S. commercial revenue can keep growing at a high speed, whether contract value can smoothly convert into revenue, whether international markets can be opened up, and how much profit margin the company can preserve after growth slows.
Palantir no longer really needs to prove it’s a good company.
The real question now is: at a valuation close to 40 times annual revenue, how good does a company have to be for it to count as a good investment?
This article is for analyzing earnings reports and business models only and does not constitute investment advice.
