There is a question that has quietly consumed every serious investment committee conversation about enterprise software this year. Not whether AI is real. Not whether companies are spending on it. The question is simpler and harder to answer: is any software company actually capturing that spending at a scale that justifies what the market is pricing in?

Monday night, Palantir gave one of the loudest answers Wall Street has heard in years.

Why Every Portfolio Manager Is Paying Attention

Palantir reported Q2 2026 results after the bell on August 3rd, and the numbers were not a modest beat. They were a statement. Revenue came in at $1.94 billion against Wall Street expectations of $1.80 billion, a 93% surge year over year. Adjusted EPS hit $0.41, well above the $0.35 consensus estimate. U.S. commercial revenue grew 149% year over year to $764 million. U.S. government revenue grew 90% year over year to $809 million, blowing past the FactSet consensus of $730.9 million. The stock surged 12% after hours.

Here is what makes this more than a headline number. Palantir has now beaten EPS expectations for nine consecutive quarters. Yet heading into this report, the stock had lost roughly 29% year to date, weighed down by persistent valuation skepticism and fears that the AI software cycle was losing momentum. That tension, strong fundamentals against a punished stock price, is exactly the kind of thing that forces investment committees to reassess their frameworks.

The full-year revenue guidance was raised to $8.15 to $8.16 billion, up sharply from prior guidance of $7.65 to $7.66 billion, implying approximately 82% year-over-year growth. U.S. commercial revenue guidance was raised to in excess of $3.42 billion, representing at least 134% growth. Adjusted free cash flow guidance was lifted to $4.5 to $4.7 billion.

CEO Alex Karp put it bluntly on CNBC: “To my knowledge, no business at our scale has even grown half this much.”

The Bull Case

The strongest version of the bull argument is not about one quarter. It is about structural positioning at a moment when AI is moving from experimentation to operations.

Palantir is not building large language models. It is building the software layer that allows governments and enterprises to actually use AI inside their own data environments, with their own rules, without handing proprietary information to a model company. That distinction matters enormously for regulated industries and national security customers.

The U.S. commercial remaining deal value, a forward indicator of future revenue, reached $6.24 billion as of Q2, more than doubling from a year ago. That is not a company benefiting from a one-time wave of curiosity contracts. That is a company with a growing backlog of committed enterprise demand. U.S. commercial revenue has now jumped 380% since 2024 when compounding is taken into account.

On the government side, the Maven Smart System’s designation as a Pentagon program of record in March 2026 is the kind of contract event that institutionalists genuinely care about. It means dedicated budget lines, formal oversight, and a defined acquisition pathway across military branches. Maven has already been used operationally to compress battlefield targeting cycles from hours to minutes. The Army’s enterprise agreement with Palantir, valued at up to $10 billion over a decade, sits underneath all of this.

The profitability story is also real. Palantir posted a Rule of 40 score of 145% in Q1 2026, and Q2 adjusted operating margins remain near 60%. This is not a company burning cash to buy growth. Adjusted free cash flow guidance of $4.5 to $4.7 billion for the full year 2026 is the kind of number that starts to make even skeptical value investors do double takes.

The Bear Case

Here is the part that sophisticated investors do not skip over, even after a quarter like this.

The valuation is extraordinary by almost any framework. Heading into earnings, the stock was trading at approximately 38 times projected 2026 revenue with a forward P/E above 100x. Even after the stock’s year-to-date pullback of nearly 30%, the embedded expectations in the price remain extreme. A quarter where growth decelerates even modestly, say from 93% to 65%, would likely cause significant multiple compression regardless of the absolute revenue number.

The government concentration risk is also worth naming. A material portion of Palantir’s revenue depends on U.S. defense and intelligence spending. That creates an unusual dependency on policy decisions, budget negotiations, and geopolitical circumstances rather than pure commercial demand cycles. The Maven contract is massive. But program-of-record status can also be renegotiated.

And the international business, while present, is not keeping pace. European commercial growth has lagged the domestic surge, which means the company’s geographic concentration in the U.S. is actually increasing, not diversifying, as the overall business scales.

There is also an analyst coverage dispersion worth noting. Northland Securities had an EPS estimate roughly 25% below consensus heading into this quarter. That kind of model disagreement signals genuine uncertainty about Palantir’s business, not just noise.

What the Evidence Actually Shows

One data point from this quarter deserves more attention than it has gotten.

U.S. commercial revenue is now approaching the size of the government business. A year ago, the gap between the two segments was roughly $120 million. In Q2 2026, U.S. government revenue was $809 million and U.S. commercial was $764 million. That gap has essentially closed. For a company that started as a pure government contractor, that is a fundamental shift in who is buying and why.

What it tells institutional investors is that Palantir’s Artificial Intelligence Platform is not just a government procurement story. Enterprise companies across healthcare, banking, manufacturing, and other sectors are signing and expanding contracts at a pace that is now generating revenue comparable to some of the largest defense software programs in the country. That changes the risk profile of the business meaningfully.

Palantir has beaten revenue and EPS estimates in each of the last nine quarters. Over the past three months heading into this report, 20 upward earnings revisions and 21 upward revenue revisions had already been submitted by sell-side analysts. The market knew the bar was rising, and Palantir still cleared it by a wide margin.

What Serious Investors Are Thinking

The most experienced investors in this space are not debating whether Palantir is growing. They are debating whether the growth is defensible, and whether what they are seeing is genuinely a new category of enterprise software or a cycle that benefits from an extraordinary moment in AI spending that will eventually moderate.

Baird has maintained a $200 price target, calling Palantir the premier AI growth asset. That view is grounded in the idea that AIP deployments are converting signed deal value into durable recurring revenue at a rate that the market has consistently underestimated. The remaining deal value of $6.24 billion supports that view. Each quarter that converts backlog into revenue at this pace is a quarter that makes the forward growth story more credible, not less.

The counterargument, held by a meaningful portion of institutional holders, is that the current valuation requires Palantir to execute flawlessly for years. One quarter of deceleration, one large contract loss, one policy shift in defense spending, and the stock faces a repricing that the math makes very hard to absorb. At 38 times forward revenue, there is almost no margin for operational error.

Both of those positions can be right simultaneously, which is why PLTR remains one of the most genuinely contested institutional positions in enterprise software.

What Investors Are Missing

Most of the coverage after Monday’s earnings focused on the top-line beat and the stock’s after-hours move. The part that deserves deeper scrutiny is what the sovereign AI theme means for Palantir’s pricing power going forward.

Karp made a pointed comment after the quarter: “Demand for AI sovereignty has now been unleashed. And Palantir is the only company that has demonstrated it can transform tokens into actual economic value.” That is not marketing language. It reflects a real competitive dynamic. Enterprises that want to deploy AI without feeding their proprietary data into a shared model have very few credible options at scale. Palantir is one of them.

Slight tangent, but it matters here. Google’s original refusal to engage with military AI applications under Project Maven, driven by employee pressure in 2018, handed Palantir a decade-long head start in defense AI infrastructure. That early positioning is now compounding in a way that is structurally difficult for late entrants to replicate quickly. You do not rebuild years of cleared personnel, classified deployment experience, and embedded agency relationships in two years.

The second overlooked angle is what happens when U.S. commercial revenue surpasses government revenue permanently. The investor base currently discounts Palantir’s commercial business relative to peers because it is newer and less proven at scale. If commercial continues to grow faster than government, that perception gap becomes a potential catalyst in its own right.

Stocks to Watch

Palantir Technologies (PLTR). The obvious one, but worth framing correctly. This is not a momentum trade. The question for institutional investors is whether the commercial backlog of $6.24 billion converts into durable revenue at a pace that justifies the current multiple over a three to five year horizon. If it does, the valuation argument eventually resolves itself through earnings growth. If commercial growth stalls at all, the stock is exposed to significant downside from its current level.

Microsoft (MSFT). Palantir’s AIP integrates with Azure and other Microsoft cloud infrastructure. As Palantir’s enterprise deployments scale, Microsoft benefits from increased cloud consumption without needing to win the AI software layer itself. Every new Palantir AIP contract that runs on Azure is a tailwind for Azure revenue, and that relationship is deepening.

Booz Allen Hamilton (BAH). The government consulting firm most likely to feel Palantir’s growth pressure and benefit from it at the same time. Booz Allen has been aggressively embedding AI into its defense consulting work, but the more embedded Palantir becomes in Pentagon operations through Maven, the more it competes for the same digital transformation budgets that Booz Allen has historically owned. Watch this relationship closely.

C3.ai (AI). The clearest competitive pressure case. C3.ai has struggled to demonstrate Palantir’s kind of revenue acceleration in enterprise AI deployment. As Palantir’s commercial backlog builds and its customer count grows, the space available for a competing enterprise AI platform shrinks. C3.ai’s next few quarters will tell investors a lot about whether this market can support multiple scaled players or whether Palantir is consolidating share.

Leidos (LDOS) and General Dynamics IT (GD). Both are large defense IT contractors that operate in the same general orbit as Palantir’s government business. As Maven’s program-of-record status expands Palantir’s footprint across military branches, some of the adjacent systems integration and data management work that traditionally flowed to larger defense IT primes could be redirected. The budget dynamics in defense AI are not zero-sum, but they are worth monitoring.

Here is the thing about a quarter like this. It does not resolve the debate. It intensifies it. Bulls walk away with more evidence that the commercial engine is real. Bears walk away knowing the stock is now more expensive than it was before the report. Both are reading the same numbers.

What the quarter actually tells you is that the AI software adoption cycle, at least inside Palantir’s customer base, is not slowing down. Whether the rest of the market catches up to that reality is a different question entirely.

Wall St. Mavens