How Palantir makes its money
Palantir Technologies is an American software company. It was co-founded in 2003 by Peter Thiel and is led by Alex Karp. The stock has been listed since September 2020. The product is harder to grasp than a chip or an online shop. Put simply, Palantir connects an organization’s scattered data sources into a usable overall picture.
It has several platforms for this. Gotham is aimed at defense ministries, intelligence services and security agencies. Foundry is the data operating system for companies. It maps a company’s data, logic and processes in a shared model, the so-called ontology. Apollo delivers the software and keeps it up to date across different environments.
The fourth platform is called AIP, short for Artificial Intelligence Platform. According to the quarterly report to the US Securities and Exchange Commission, AIP connects large language models with customers’ data and processes. Companies use it to build AI agents that take on specific tasks, for example in logistics, manufacturing or insurance. It is precisely this combination of language model and operational data that is currently driving demand.
Palantir serves two customer groups. In the second quarter of 2026, $990 million came from government customers and $945 million from commercial customers. So around half of the business depends on government contracts, mainly from the US.
Two stories, one stock
With Palantir, two narratives run in parallel. The first is about the business: a software provider is growing faster than almost any other large technology company and making high profits in the process. The second is about the price: the stock is among the most highly valued in the world.
Both narratives are true. Anyone who tells only one of them distorts the picture. That is why I examine the two levels separately. First the business, with its figures and drivers. Then the valuation, with the question of how much of the future the share price already anticipates. A strong company and an expensive stock can coexist for a long time. But they can also reconcile abruptly, upwards as well as downwards.
AIP drives growth
The latest figures are exceptional. According to the press release on the second quarter of 2026, revenue rose by 93% to $1.94 billion. The US commercial business grew by 149% to $764 million. The US government business rose by 90% to $809 million. Palantir closed 220 deals with a value of at least one million US dollars.
The company is making real money in the process. The quarterly report shows a net profit of around $1.07 billion for the quarter. For the full year 2026, management expects revenue of around $8.15 billion, an increase of 82%.
The driver behind these figures is demand for AIP. Many companies have experimented with chatbots and now want to bring AI into their core processes. For this, Palantir sells short pilot projects, so-called bootcamps, which often lead to large contracts. The gap is striking, however: outside the US, the commercial business recently grew by only around 26%, according to the quarterly report. So far, Palantir is above all an American growth story.
What the valuation means
Now to the second narrative. On August 10, 2026, the market capitalization was around $429 billion according to Stockanalysis. This compares with $6.16 billion in revenue and $3.02 billion in profit over the last twelve months. That results in a price-to-earnings ratio of around 147. The market value corresponds to about seventy times the twelve-month revenue.
For comparison: established software groups are often valued at five to fifteen times their revenue. Palantir costs a multiple of that, even measured against its own forecast for 2026.
What does that mean in concrete terms? A valuation like this is not a snapshot of today’s business. It is a bet that revenue and profit will keep growing very fast for many more years. If you work the expectation backwards, Palantir has to deliver almost flawlessly over a long period. If growth slows earlier than expected, the share price can fall significantly even though the company keeps growing. Conversely, the stock can stay highly valued as long as the figures beat expectations. Neither of these is a forecast. It is simply the mechanics of high multiples.
Three questions before a possible purchase
The first question concerns the business. Is growth accelerating further, or is it leveling off? With Palantir, I look in particular at the US commercial business, order intake and whether international customers follow suit. A single record quarter says less than the trend over several quarters.
The second question concerns the trend. Is the stock trading above its long-term moving averages, and are these lines rising? This is not a signal to act. It only shows me whether a position would put me against the prevailing direction.
The third question concerns momentum. After strong moves, a stock can be overheated in the short term. The Relative Strength Index helps with this assessment. Especially with a stock valued like this, I check whether buying in would be rash or whether the situation has calmed down.
The main risks
The valuation leaves little room. A price-to-earnings ratio of around 147 prices in years of high growth. Even a slight disappointment can trigger a strong price reaction without the business getting any worse.
The government business is concentrated. Around half of revenue comes from government customers, mostly in the US. Budget disputes, elections or new priorities at individual ministries can delay or shrink large contracts.
Political closeness is a double-edged sword. Palantir benefits from close ties to the US government and the military. The same closeness makes the company vulnerable when political majorities shift. In Europe, its work for security agencies also regularly sparks debates about data protection and dependence.
Stock-based compensation is high. According to the quarterly report, it cost around $265 million in the second quarter of 2026, 66% more than a year earlier. The number of shares rose from 2.34 billion at the end of 2024 to 2.40 billion in mid-2026. Every new share dilutes your share of the profit.
The international business is lagging behind. More than 80% of revenue is generated in the US. If the rest of the world remains hesitant, the growth story depends on a single market.
My conclusion on Palantir stock
Palantir is currently delivering what many AI companies only promise: measurable growth, real profits and software that runs productively in government agencies and large companies. Its position at the intersection of data, AI and security is rare and hard to copy.
The stock is still not a simple case. The market pays for this business with a valuation that assumes many years of perfect execution. So the central question is not whether Palantir is a good company. The question is how much of that quality is already reflected in the share price.
For my own assessment, I therefore separate three levels: business development, the long-term trend and short-term momentum. This sequence does not replace a personal investment decision. But it ensures that enthusiasm for the technology and respect for the price stay in the same picture.