Palantir (PLTR.US) Q3 Conference Call: Commercial and Government Revenue Both Break $1 Billion, Sovereign AI Revolution Has Just Begun
The Zhitong Finance App learned that Palantir Technologies (PLTR.US) announced results for the third quarter of 2026: revenue increased 96% year-on-year, 20% month-on-month, and quarterly revenue exceeded 2 billion US dollars for the first time. The company said demand for its “Sovereign AI” (Sovereign AI) products drove particularly strong growth in the US commercial business, which surpassed US government business revenue for the first time this quarter.
Total revenue for the third quarter was US$2,313 million, of which revenue from the US was US$1.937 billion, up 119 percent year over year. Chief Revenue Officer and Chief Legal Officer Ryan Taylor said that US revenue accounts for 84% of the company's total revenue.
Commercial business revenue surpassed $1 billion
Commercial revenue increased 113% year over year and 24% month over month to reach US$1,168 billion, marking the first time that the business segment surpassed $1 billion in a single quarter. US commercial revenue increased 148% year over year and 29% month over month to reach 984 million US dollars.
Chief Financial Officer Dave Glazer said that US commercial revenue increased by US$587 million compared to the third quarter of 2025. The company's total contract value (TCV) bookings for the US commercial business reached US$1,463 billion this quarter, an increase of 11% over the previous year. In the past 12 months, US commercial business TCV bookings reached US$6.114 billion, an increase of 62% over the previous 12 months.
Palantir said the number of its US commercial customers reached 686, up 29% year over year and 5% month on month. The total value of remaining contracts in the US commercial business increased 79% year over year. International commercial business revenue increased 21% year over year to reach US$184 million.
Taylor cites several customer expansion cases, including an industrial materials company turning a nearly $3 million pilot project into an agreement with an annual contract value (ACV) of $29 million. A global sporting goods company turned a $1.6 million pilot project into an ACV agreement of approximately $21 million, while a global appliance company expanded the $1 million pilot project into a $15 million ACV agreement covering commercial, financial, and supply chain workflows.
The company said that revenue from strategic commercial contracts in the third quarter was approximately US$400,000, accounting for less than 0.02% of total revenue. The company expects revenue from these contracts to be less than $500,000 in the fourth quarter and will no longer generate revenue thereafter.
The government business sector also surpassed $1 billion
Government business revenue increased 81% year over year and 16% month on month to reach US$1,146 billion, also breaking the US$1 billion mark for the first time. US government business revenue increased 96% year over year and 18% month on month to reach 953 million US dollars; international government business revenue increased 31% to reach 192 million US dollars.
Taylor said the growth of Palantir's government business reflects its work on civil and defense projects, including the Maven project. He pointed out that the US Army awarded the company a prime contractor agreement in the third quarter to produce eight TITAN systems.
Chief Technology Officer Shyam Sankar said that Maven currently has more than 100,000 users and has been used by various military branches, combat headquarters, and many combat support agencies. He also highlighted the company's “Disconnected Ontology Resilient Architecture” (Disconnected Ontology Resilient Architecture, DORA for short), describing it as an embedded computing device designed to bring Maven capabilities into a disconnected combat environment.
In terms of the company as a whole, Palantir's TCV bookings for the quarter reached US$2,918 million, an increase of 6% over the previous year. The net dollar retention rate was 165 percent, an increase of 800 basis points over the previous quarter. The total value of remaining contracts reached US$13.4 billion, up 55% year on year; remaining performance obligations were US$4.9 billion, up 88% year on year.
Profitability and cash flow continue to expand
Palantir reported GAAP operating profit of US$1,085 million, with a profit margin of 47%; GAAP net profit of US$1,128 million, with a profit margin of 49%. GAAP and adjusted earnings per share were $0.44.
The adjusted operating profit was US$1,427 million, and the adjusted operating margin was 62%. The adjusted gross margin was 85%. The cash flow from the company's operating activities was US$1,255 million, and the adjusted free cash flow was US$1,216 million, with corresponding profit margins of 52% and 53%, respectively.
By the end of the quarter, Palantir held a total of $10.4 billion in cash equivalents and US Treasury bonds. Equity incentives cost $319 million, and employers' payroll tax costs associated with equity are $23 million.
According to Glazer, adjusted expenses increased 53% year over year and 20% month-on-month to reach 887 million US dollars, mainly due to the company's investment in AI platforms, R&D and technical personnel recruitment.
The company raised its annual results guidelines
For the fourth quarter, Palantir expects revenue between US$2,649 million and US$2,653 million, and adjusted operating profit of between US$1,640 million and US$1,644 million.
For the full year of 2026, the company raised its revenue guidance to the range of US$8.530 billion to US$8.534 billion. The median guideline represents a year-on-year increase of 91%, which is 9 percentage points higher than the company's previous annual revenue growth guide.
US commercial revenue for the full year is expected to exceed US$3.559 billion, representing an increase of at least 143%; adjusted operating profit is expected to be between US$5.245 billion and US$5.249 billion; adjusted free cash flow is expected to be between US$4.8 billion and US$5 billion; and the company continues to expect GAAP operating profit and net profit for every quarter in 2026.
In a conference call, CEO Alex Karp said that customers increasingly want to maintain control over their operational data and AI models, which Palantir refers to as “sovereignty.” Sankar said the core of the company's approach is to deploy AI in enterprise workflows rather than just focusing on the productivity of individual knowledge workers, with the goal of generating measurable operational returns.
Full summary content
Ana Soro (Finance Team)
Good afternoon. I'm Ana Soro from the Palantir Technologies' finance team, and I welcome everyone to our third quarter 2026 earnings conference call. We'll be discussing the results announced in today's post-close press release, which was posted on our investor relations website. At this meeting, we will publish statements about the company's business that may be considered forward-looking statements under applicable securities laws, including statements relating to our fourth quarter and fiscal year 2026 results, management's expectations for future financial and operating results (which do not constitute official guidance unless otherwise specified in the attached written materials), and other statements about our plans, prospects, expectations, goals and vision.
These statements are not promises or guarantees, and are subject to risks and uncertainties that may cause actual results to differ materially from those statements. Information on these risks can be found in our earnings press release issued after today's close and in our filings with the U.S. Securities and Exchange Commission. We are under no obligation to update forward-looking statements except as required by law. Additionally, in today's session, we'll be referring to some adjusted financial metrics. These non-GAAP financial measures should be used as a supplement to GAAP measures, not as a substitute for or in isolation from GAAP measures. More information about these non-GAAP measures, including the reconciliation of non-GAAP measures with comparable GAAP measures, is included in the press release and investor presentation materials we are providing today. Our press releases, investor presentations, and other financial materials are available on our Investor Relations website.
In this session, we'll be referring to various growth rates as we discuss the business. Unless otherwise specified, these growth rates reflect year-over-year comparisons. I was also attending today with CEO Alex Karp, Chief Technology Officer Shyam Sankar, Chief Financial Officer Dave Glazer, and Chief Revenue Officer and Chief Legal Officer Ryan Taylor. Now I'm going to leave the meeting to Ryan.
Ryan Taylor (Chief Revenue Officer and Chief Legal Officer)
Humans are facing the world's historic crossroads of how to choose to deal with AI; and in the process, our pioneering technology has always led trends and redefined the industry. We've been shaping the AI revolution from the beginning, and although the sovereign AI revolution has only just begun, we've seen how it's changing the size of the market. We achieved a 96% year-over-year increase in revenue and a 20% month-on-month increase, which is the highest delivery growth rate in our history as a listed company. These results are a continuing financial reflection of the transformative impact we have had on our clients, and this is reflected in the depth of our customer relationships. In the third quarter, the average revenue of our top 20 customers over the past 12 months increased 76% year over year to reach US$146 million per customer.
Our US business is running at a high speed. We achieved US revenue of $1.9 billion, up 119 percent year over year and 23 percent month on month, accounting for 84% of our total revenue. This was mainly led by our US commercial business, which grew 148% year over year, 29% month over month, and surpassed our US government business revenue for the first time this quarter. In the US commercial business, the average revenue per customer over the past 12 months increased 88% year over year and 20% month over month.
The US remains at the center of the AI revolution. I'm not going to repeat what I said last quarter because this is self-evident in our unrivaled results. Two years ago, we led the trend with the “commoditized perception” theory, but now the market is only beginning to recognize the full power of sovereign AI. This is pervasive in my conversations with customers, and it's undeniable in the results they share. OptumServe's chief operating officer emphasized that sovereignty doesn't mean we have to structure everything ourselves. But it does mean protecting our differentiating strengths while maintaining the ability to evolve, adapt, and expand solutions as customer needs change—and customer needs change frequently.
That means knowing where the right partner can help us move faster. L3Harris notes that by fine-tuning the open source model on its own data, a sovereignty model surpassed the cutting-edge model it had previously used in less than 48 hours and reduced costs by 95% in initial results. NVIDIA reports that by evaluating and post-training its open model, it has made its smallest and most efficient model surpass its largest model by a large margin, and can now retrain the model in just a few hours on a single GPU, enabling them to retrain at any time as long as they have data to improve decision intelligence.
They said, “We captured our own alpha and trained it into a model. We can't start this process until we've built a workflow in Palantir.” Our clients are rapidly choosing Palantir and its sovereign AI to transform their organizations. An industrial materials company piloted a nearly $3 million project in the second quarter of 2026, which then turned into an annual contract worth $29 million. Last quarter, a global sports company turned a $1.6 million pilot launched in the first quarter of 2026 into a deal with an annual contract value of approximately $21 million. Last quarter, a global home appliance company converted the $1 million pilot launched in the second quarter of 2026 into a $3 million annual contract value transaction in the same quarter, then expanded nearly fivefold in the third quarter to reach an annual contract value of $15 million, covering commercial, financial, and supply chain workflows.
Following an agent camp in the second quarter of 2026, a clothing company signed a $2.5 million pilot, which was then converted into a deal worth $11.5 million in annual contract value last quarter. These promises are our delivery of downstream results with measurable impact to our customers. With Palantir's sovereign AI, McCarthy Building Companies will save 30,000 hours per year, using post-trained Nemotron to help translate complex project requirements into contract language to allocate risk and clearly define subcontractor obligations. Tampa General is deploying sovereign AI to keep sensitive data and model processing locally while enforcing its own risk and privacy standards for LLM traffic, and to support Tampa General's desire to achieve this result using fine-tuned and post-trained Nemotron.
Together, we improved model performance by 7% and reduced latency by a factor of 22. Shifting to our US government business, revenue increased 96% year over year and 18% month over month, demonstrating continued extraordinary influence in the civil and defense sectors. Using MAVEN as our foundation, in the third quarter, the Army granted Palantir a master agreement to produce eight TITAN systems; last quarter, the FAA emphasized the impact of our collaboration, pointing out that the shift from distributed data to embedded entities is the key to changing the rules of the game. This is modernization in the form of operations, changing the way security work is done. They point out that what they can do now without Palantir would require hundreds of analysts.
Looking back at our performance, I see that this rapid pace continues unabated. I'm more excited than ever because of the ambition I've heard from customers and their demand for sovereign AI — we should remember that this revolution is only a quarter past. Despite these remarkable results, they are just the beginning of sovereign AI. Now I'm going to leave the meeting to Shyam.
Shyam Sankar (Chief Technology Officer)
Thanks, Ryan. Palantir is obsessed with turning tokens into economic value. Obviously, the vast majority of economically valuable problems that LLM can solve can be solved with intelligence at or below the cutting edge level. So what's stopping this from happening? At one end are model makers, who can't imagine solving any problems other than by improving the model, so they haven't even tried it. At the other end is the hard work of diffusion, which sounds passive, but in reality it is by no means the case.
All value comes from the elegant integration of AI software and humans. This requires not a model harness, but something closer to an enterprise harness. And that's AIP. Most profoundly, the laboratories mistakenly believed that the atomic unit of an enterprise is a personal knowledge worker. That's not the case. The basic unit of an enterprise is a workflow. You can make any group of workers more productive without generating any real ROI.
For something, like programming, a knowledge worker is a workflow, so there's a perfect product market fit. But for the vast majority of businesses, that's not the case. In view of this, AIP can enable enterprise workflows and introduce AI in a way that generates objective, quantifiable value, which is extremely valuable. For example, page 209 of Anthropic S-1 talks about AIG's automation of complex commercial insurance underwriting processes. The entire workflow and elegant applications of AI software and humans run in AIP and are implemented by PalantiR and AIG. We are a diffusion layer, not only a choice, but also a necessity.
Among our customers, those most successful in deriving value from AI are showing a clear pattern. They have huge ambitions. Their goal isn't something as banal as lowering operating expenses. Instead, it's market-driven. The opposite is also true. The most struggling customers are narrowly focused on reducing costs. AI is a new kind of intelligence. Just trying to cram it in to replace humans completely missed this era. The mistake of confusing human intelligence with machine intelligence will lock you in on an improperly chosen goal that limits your ambitions and remains out of reach.
Instead, we're overcoming all the work companies know they should do, but have historically lacked the financial resources or ability to do: an agent, Swarm, observes every patient in the ICU ward to help clinicians detect potential signs of sepsis earlier and improve patient outcomes. Each wellhead in the Permian Basin was equipped with an agent swarm to maximize production. Each of our clients is equipped with an agent swarm to drive growth. It's not about doing the same thing at a lower cost.
It's about defining productive work that surpasses yesterday's frontiers, and capturing and compounding the new economic value it creates, and does it in the form of sovereign AI. Examples are everywhere. It is now clear that post-training on your specific operational data and ontology can achieve superior performance and protect your alpha with weights you control yourself. Yes, it's cheaper, but the dominant variable is: when machine intelligence is gracefully integrated into your business and our products, more use cases can actually reach the threshold of viability to generate economic value.
Investments in AIP to create a closed loop model factory are making it all simple and fast. A specific post-training model is a commodity. The compound value lies in the post-training pipeline, which continuously takes input from your operating exhausts, which are captured in the body to drive automated post-training runs, model checkpoints, evaluations, and deployment across an open weighting ecosystem yesterday, today, and tomorrow. Our new feature Auto Research and continued investment in AIP Evolve are bringing us closer to an automated cycle to continuously build, deploy, and evaluate intelligence based on ROI in AIP and sovereign AI.
Pareto's Best Frontiers is no longer an unattainable academic ambition. It's a simple function call that we can automatically call on your behalf. In a recent PAYCOM exercise, we deployed a 30 billion parameter post-training model to a computer limited by space, weight, power, and cooling to prepare for disconnected combat. Its specific mission performance was superior to GPT 5.6 — a 3 trillion parameter model. L3Harris achieved 95% cost reduction through sovereign AI. McCarthy will save 30,000 hours per year on subcontractor drafting, achieve superior RFI classification and subcontractor drafting performance on a post-trained Nemotron 4 billion parameter model, and once again, the real value lies in the significant increase in the number of demand events they can handle.
Now, Maven continues to deliver results to the coalition forces. With over 100,000 users, Maven can now be used not only by combat commands, but also by all military branches and many combat support agencies. We just finished a Maven hackathon. A total of 46 coalition teams participated. Our community of 25,000 builders has built so many amazing projects. The two winning teams were Army teams, one built edge planning and simulation capabilities for force protection, and the second place built an incredible platform to analyze drone intrusions to drive troop protection insights.
We continue to feel the demand for Maven extends to the tactical edge, in disconnected environments on shore and at sea. We have had incredible success with DORA (Disconnected Ontology Resilient Architecture), including in the Army's Jailbreak effort. DORA is an embedded computing device that brings Maven's capabilities to disconnected combat. Using AIFD, service personnel can build, read, write integrations with legacy sensors and effectors, and bring them online in a matter of hours without assistance at all.
Next, I'll leave the meeting to Dave.
Dave Glazer (Chief Financial Officer)
Thanks, Shyam. We had a great third quarter. We achieved the highest reported year-over-year revenue growth rate of 96% in history and the highest reported month-on-month revenue growth rate of 20% in history. Quarterly revenue surpassed the $2 billion mark for the first time, and we also surpassed multiple billion dollar milestones in commercial revenue, government revenue, and GAAP operating profit. Revenue from our US business accelerated again, growing 119 percent year over year and 23 percent month over month.
Our US commercial business grew 148% year over year and accelerated to 29% month over month; our US government business grew 96% year over year and 18% month over month. Demand for our sovereign AI products is accelerating our US commercial momentum, helping to make this quarter the first quarter in history where US commercial revenue surpassed US government revenue. This is a key milestone in our company's history. In addition, we raised our full-year US commercial revenue guidance to more than $3,559 million, representing a growth rate of at least 143 percent. We also raised the midpoint of our 2026 full-year revenue guidance to US$8.532 billion, representing a 91% year-on-year increase, which is 9 percentage points higher than the previous quarter's 2026 full-year revenue guidance.
Shifting to our global top-line results, third-quarter revenue increased 96% year over year and 20% month over month to reach US$2,313 billion. In the third quarter, US revenue increased 119 percent year over year and 23 percent month on month, reaching US$1.937 billion. Revenue from our biggest customers continues to grow. In the third quarter, the revenue of our top 20 customers over the past 12 months increased 76% year over year to reach US$146 million per customer. Now on to our commercial section. Commercial revenue for the third quarter increased 113% year over year and 24% month on month to reach US$1,168 billion, breaking the $1 billion mark for the first time. We completed $1,652 billion in commercial TCV bookings in the third quarter, up 17% year over year. Our AI platform continues to dominate the US market as the only real choice for putting LLM into operation, while still giving customers full control over the data, logic, actions, and security of their businesses.
In the third quarter, US commercial revenue increased 148% year over year, and accelerated to 29% month-on-month, totaling US$984 million. US commercial revenue for the third quarter of 2025 was $397 million, which represents an increase of $587 million in just one year. We completed US$1,463 billion in US commercial TCV bookings, an increase of 11% year over year. Over the past 12 months, we have completed $6.114 billion in US commercial TCV bookings, an increase of 62% over the previous 12 months, highlighting the accelerated demand for AI — AI creates real operational value while shifting power from the lab back to your business.
The total remaining transaction value of our US commercial business increased 79% year over year and 4% month over month. We believe sovereign AI is the latest and greatest technological revolution of our generation. We signed multiple eight-digit deals this quarter, with shorter terms than usual as customers quickly adopted initial sovereign AI capabilities in their businesses. It's important to remember that the sovereign AI revolution is just over a quarter. As we continue to deliver value through sovereign AI over time, we expect both the duration and size of these contracts to increase.
The number of our US commercial customers grew to 686, up 29% year over year and 5% month on month. International commercial revenue for the third quarter increased 21% year over year and 1% month on month to reach US$184 million. Third-quarter revenue from strategic commercial contracts was approximately US$400,000, accounting for less than 0.02% of total revenue. We continue to expect revenue from these contracts to be less than $500,000 in the fourth quarter, and we expect no more revenue from these contracts after the fourth quarter of this year.
Shifting to our government sector, government revenue in the third quarter increased 81% year over year and 16% month on month to reach US$1,146 billion, breaking the $1 billion mark for the first time. In the third quarter, US government revenue increased 96% year over year and 18% month on month, reaching US$953 million. This growth is driven by continued execution of existing projects and new awards, reflecting the government's growing demand for our AI platform. In the third quarter, international government revenue increased 31% year over year and 6% month on month, reaching US$192 million.
We completed TCV bookings of $2,918 million, an increase of 6% year over year. The net dollar retention rate was 165 percent, an increase of 800 basis points over the previous quarter. At the end of the third quarter, our total remaining transaction value was US$13.4 billion, up 55% year on year and 2% month on month; remaining performance obligations were US$4.9 billion, up 88% year on year and flat month on month. As a reminder, RPO mainly consists of our commercial business because it doesn't take into account contracts with an initial period of less than 12 months and contractual obligations beyond convenient termination clauses, which are mostly common in our government business.
Moving to profit margins and expenses, the adjusted gross profit margin (excluding stock compensation expenses) was 85% for the quarter. Adjusted operating profit (excluding stock compensation expenses and related employer payroll taxes) for the third quarter was US$1,427 million, representing an adjusted operating margin of 62%. The adjusted expenses for the third quarter were US$887 million, up 20% month-on-month and 53% year-on-year, mainly driven by our continued investment in AI platforms and technology recruitment, as we welcomed the latest new graduates.
We remain committed to investing in the most elite technical talents and research and development of our product pipeline and sovereign AI work while achieving the goal of continued GAAP profitability. GAAP operating profit for the third quarter was US$1,085 million, representing a profit margin of 47%, and broke the $1 billion mark for the first time. GAAP net profit for the third quarter was $1,128 million, representing a 49% profit margin. Net profit has increased nearly eight times over the past 24 months, rising from $144 million in the third quarter of 2024.
Third-quarter stock compensation expenses were US$319 million, and equity-related employer payroll tax expenses were US$23 million. GAAP earnings per share for the third quarter were 44 cents. Third-quarter adjusted earnings per share were 44 cents. Furthermore, the sum of our revenue growth and adjusted operating margin accelerated to 158% in the third quarter, an increase of 3 percentage points over the previous quarter. This is the 13th consecutive quarter in which our Rule of 40 score has increased.
Turning to our cash flow, we generated $1.25 billion in operating cash flow and $1,216 billion in adjusted free cash flow in the third quarter, representing 52% and 53% profit margins, respectively. At the end of the quarter, we had $10.4 billion in cash, cash equivalents, and US Treasury securities. Turning now to our outlook for the fourth quarter of 2026, we expect revenue of between $2,649 million and $2,653 million and adjusted operating profit of between $1,640 million and $1,644 billion. For the full year of 2026, we raised our revenue guidance to between $8.530 billion and $8.534 billion. We raised the US commercial revenue guide to more than $3,559 million, representing a growth rate of at least 143 percent. We raised the adjusted operating profit guidance to between $5.245 billion and $5.249 billion.
We raised our adjusted free cash flow guidance to between $4.8 billion and $5 billion, and we continue to expect GAAP operating profit and net profit every quarter this year. Next, I'll leave the meeting to Alex to talk about a few words, and Ana will start the Q&A session.
Alex Karp (CEO)
Welcome to our earnings conference call. You know, I think for many, many years, our earnings call was actually only about our profits. It is now the focus of technology, society, and general trends. I think the people watching this conference — many, if not most — are the most watched earnings conference calls in the world — people are here to find out what's really going on at the bottom. Since AI and sovereign AI are transforming the universe, in our case, they are bringing huge benefits, and as reflected in our numbers, people really want to understand what is actually happening in the world politically, sociologically, anthropologically, and of course financially.
So, given the demand and interest in Palantir, and the moment that has become the real focus of understanding reality, frankly speaking, many investors watching this conference are looking at what really happened to their investments, or even mostly not at Palantir. So there's a bigger question: What does it mean to have a business built around what we believe in and still believe is the most important mission in the world — supporting our fighters and bringing them home? Even a year or two ago, even those involved within Palantir viewed the core values of our business from an economic perspective, as Glazer would. Frankly, in a very helpful way, he's a power of greed.
You know what Glazer thinks about our business is, can you talk more about commercial business? Over the past year, past few years, and this quarter, our commercial business has surpassed government business. In fact, considering our overall business growth of 96%, the value of this market probably justifies our current valuation. The behemoth of American business — I mean, these numbers far exceed what you can — I mean beyond what even the most optimistic people can think a business like ours can do, while maintaining a completely unique way of going public.
We have unique products. We have a unique way to go public. We're the leaders of a revolution in front of us, but none of us can really name the revolution of sovereignty, which means people need to understand how AI creates value, how you create value from tokens, how you maintain the value of your alpha, how it doesn't transfer to anyone else, and how you can extend it. Of course, this revolution is taking place in America. And then these unbounded results, we are the metaphor, really showing you that the future is becoming the present, and now is the present. You can see it in Europe, where I spent half my life. We should avoid talking about it at this point because of what is happening and simply stomping on them doesn't help. But you can see it all over America.
You see the reaction to AI, you see political polarization — because all the value in the market is actually shifting to those carriers, in this case Palantir and a few others — Anthropic, OpenAI, us, others — where you can actually claim some value creation exists. Again, this is extreme. A company that was considered the icing on the cake of our business until three years ago is now a dominant part of our business. That business grew 29% month-on-month. It has grown by almost 150% and has overtaken our US growth figure by 100%. These are just that. And we're doing this while attacking the market the way we attack it. Very few sales people can't understand how we captured this giant unless you see it as a new and different moment.
So when people watch this—what's happening in the market, what's happening in our society—it's true that scripts, products, and marketing methods don't work in most cases, but what works just suddenly comes to life. Also, you know, our ambition to take over the entire market is a bit crazy, and it doesn't seem that crazy right now. Also, you know, we believe that among this company and many viewers, we have not only financial reasons to take over the entire market; Glazer covets this. In terms of our mission, we believe we have a mission reason to take over the market.
The reason for that mission is, we support your alpha, and we support your sovereignty. We've been doing this for over 20 years. We support the fighters. We're excited to be able to support the fighters even more because we've made so much money. But we believe we have a moral responsibility to be the dominant player in our industry, broadly defined to include all AI applications and the use of open and closed models, particularly training open weighted models.
That moral responsibility drives every part of our business. The interesting thing about the internal dynamics of our business is that it's not just the sum of the parts of our business that are stronger — it's obvious. 96% increase. Check out our Rule of 40, which is 158. It's just jaw-dropping. But every part of our business, every separate part, the parts that people in the company are running, those parts are stronger than before, every part. That's what you see in these numbers.
Also, for most people watching, it's arguably what you see in the wider community. This move from the future to the present is why many classic business methods — once thought they would fail within five years — don't work. You have to do it now and now. And this company, Palantir, and everyone involved and everyone who watched and stood by us — lots of people — are part of that movement.
That's why we're so genuinely happy about the future and, quite frankly, optimistic, especially in America.
Q & A session
Ana Soro (Finance Team)
Thanks, Alex. Our first question came from Dan Ives at Wedbush. Dan, please turn on your camera and you'll be prompted to unmute the line.
Dan Ives (Analyst)
Yes, thank you. It was clearly another extraordinary quarter. So my question is about sovereign AI bootcamps. I mean, I've been to a few. They really seem to be becoming very important in converting customers. Can you walk us through, say, Alex, you've talked to so many clients, they came to a training camp — just trying to understand their AI path — and then they actually became customers when they left? Can you walk us around from your perspective?
Alex Karp (CEO)
Um, you know, I mean, what really changed on the front line was that there used to be a long process of reaching out to people, like talking to us, then you did a pilot, then you tried to get a long-term pilot, and then you locked it down. Because just like you locked in, hire all these sales people so that the market believes you'll have sales or something. Now it's more, you know, people. It's very heterogeneous. So there are people -- basically everyone is on this spectrum: what I have right now doesn't work, I've heard that yours works.
This is usually the starting point. They are also very convinced and are quite tired of what they believe. The reality is that my IP is in jeopardy and I'm not getting any value. Just like. That's when we started talking about sovereignty. This used to be a thing of the past. It's something they feel but don't know how to express. Now it's something they've expressed, and they've moved from anger to action. Like, you know, like they went from thinking, hey, my marriage wasn't good, to I went to see a divorce.
Here they are, and now what is very special is that they are surrounded by people who have gone through the same journey. The timeline from interaction to when they bought a pilot or went into production completely changed. So what Ryan and I saw on the front line of the business was faster interaction, faster deployment, and a shorter contract period, but then we were able to expand the contract significantly. Because once you go in and show them you can, for example, train open weight models in our ontology, those models are valuable, you have the ability to surpass the cutting edge without losing alpha, and you're hooked. What we have to do internally is figure out how to reconfigure our company resources to focus on those people because we don't even have enough bandwidth to do that.
So. But that's because, like, every time we go into a conference, just like in the past, we -- we go into a conference, and people are like, warm leads become hot leads. It's like getting into bed on the first date, you know? You know, maybe this analogy isn't appropriate, but it's very different. Like we—let's say it—we've gone from being the most attractive person in physics to being the captain of a rugby team.
I mean, not that I've done that, but it's like what's happening, a completely different situation. Then it's different on our side because we know we can deliver use cases that others can't. So, just like today, we're going through use cases one by one, and a super -- like one of the most important companies, I'm going to see them, basically like I want to -- rather than saying, hey, just, let's get started -- you have this critical issue in your business.
We knew we could deliver this within a few weeks. This will change the trajectory of your business. In a weird way, this is closer to what we first did in the US government, and we were like, hey, let's get started. We know we can prove it on PG, but it's like on a much larger scale. So instead of $250,000, we started at $25 million. Honestly, the overall feeling is that if this actually works, it's worth infinitely.
And it'll work. Frankly, I can't see how other products work because it's both a product and an art. It's not. That's why these numbers are so exaggerated. It's not some kind of static — when Shyam is on the front line doing implementations or Ryan is building these. I mean, Ryan was so busy in the last few days of the last quarter that he couldn't even get out of the car. We're just like that. We're like, we're just about to. I was about to deliver a bottle of wine so he could, you know, handle his natural needs.
Just like. Just like. Frontline is just that, like. I didn't even have one. I'm not exaggerating. Like, I was like, Ryan was on his way to work and he had to stop by the side. He drives a nice car. I mean, look at that. But, you know, this is. You know, but he really had to stop by the side. That's what's happening. But again, I know that the main focus of your interest is finance, but there's another side. The other stuff doesn't work. That's part of the reason. Not only is our stuff unique. There is a strong feeling in the market that these other methods — fake deployment companies, original token usage, lightweight companies that resell tokens, software companies that resell tokens and count them into revenue — people are just tired of it. Last but not least, the tastes of the market are changing.
Like, people are really going to love something that works. I mean, look at the Department of Defense's success. You know, it's like, that's it. We changed our tastes. This is what is happening in American business. That's what explains these numbers. This is it. If you--like, someone watching this, you should come. You should come to our one, you know, AIP Sovereignty. You know, it's kind of like a rock concert.
Ana Soro (Finance Team)
Thank you. Our next question comes from D.A. Davidson's Gil. Gil, please turn on your camera and you'll be prompted to unmute the line.
Gil Luria (Analyst)
Thank you. Good afternoon. I'd like to continue your conversation with Dave Dan. You talked to him about a type of CEO, the kind of person who came to you with a sense of sovereign urgency. But as reading Anthropic S-1 tells us, roughly half of their customers work directly with Anthropic. So when you have a CEO, that means a lot of CEOs haven't received information, and as you said, you're in the first game of this conversation.
When you meet a CEO who's still unconvinced, what analogies and analogies can you use to help them understand the importance of this layer of diffusion? The choreography comes from you, not directly from the lab. What examples can you provide to help them understand the risks?
Alex Karp (CEO)
Your question has two parts, and part is about that person who isn't convinced or isn't convinced, then what do we do. But strangely enough, we really can't meet the demand. So it's a market that's growing so fast. We don't need people who aren't convinced. We really don't need it. BTW, and again, I say this every time. Dario is someone I get along very well with. I think he's one of the most important business figures of my generation, if not generations.
They're also selling our products because people interact with them. Again, some people want that approach. But a lot of people are very worried about their alpha being transferred to them. And the results, or even the results they're talking about, are often already in our application layer. We've worked very effectively with Anthropic at the Department of Defense, AIG, and everywhere. But again, what they're doing is mainly expanding the size of the market.
What we do is serve a market that has become real. We're actually in a symbiotic relationship with them commercially because they're growing a market that's probably much larger than it is now. We're taking advantage of that market where delivery really is needed, and that includes delivery. Now, if there's someone we don't. We are really persuading in the government sector. Like, we've always been on the front lines, some agencies don't want to adopt it, and we know that if they don't, fighters will die. So let's try to convince them. In the business world, our real approach is, you know, hey, enjoy the one you have now until you. You found out it wasn't the one you wanted. You know, we try to compete and take their customers very, very rarely.
This really didn't happen. In turn, you know, you know, once you get on Palantir, you won't go back. This is it. That's it. Moreover. And you see that in our numbers, ie. So this is it. It's a very different thing. I don't know when we'll be competing with each other. But at this point, in the US, our market is unsatisfying. It is. And it can't be satisfied, and it's growing, growing, growing. You know, there are two ways to look at our numbers.
One is that we simply excel at what we do. The other is that we are very unique and the market explodes all around us. I prefer the latter one. We are a creative colony. We're not perfect. But the market is exploding all around us, and they're helping it explode. The only area where we do anything that looks like competition is where we spend a lot of time educating our customers, partners — and probably never partners — about the risks of using a closed model without an application layer. It's one thing we're really active in. So we're not really saying, hey, you would, you know, value creators know we have this. Of course it's an industry. I think it's actually a little less on Wall Street. Honestly.
I think on Wall Street, I'm not saying you, but I think what really happened on Wall Street is that everyone invested in these companies very early. Bless you guys. It's hard to see reality when you're making money on something that doesn't actually work the way you think it should. But the reality is, in the industry, people realise that when it becomes real, you need us, and that market is much bigger than we can serve.
operator
Thank you. Our last question came from Brad from Deutsche Bank. Brad, please turn on your camera and you'll be prompted to unmute the line.
Brad Zelnick (analyst)
That's great. Thanks so much for getting me involved. Congratulations on another rocketship-like quarter. Alex, in your shareholder letter, you stated that you saw Palantir as a pioneer of a real and viable anti-war movement. Can you expand on what this means and what it might mean for your government business?
Alex Karp (CEO)
Well, you know, one of them. I'll let Shyam talk about this. You know, I'll go back to the beginning of our Palantir PG. You know, we, we built products to protect civil liberties and stop terrorism for a simple reason: you don't want to be just one of them. If you don't stop terrorism, you'll get a radical edge on the right. If you stop terrorism when you have no rights. None of us wanted that. OK. You could say it's an anti-terrorism product, or you could say it's a social cohesion product.
It's this product because ours was built to avoid war. It means the enemy knows they're going to suffer so much that it's actually not worth waging war. It's also about creating social cohesion, because we want to bring back our troops healthier, safer, and unharmed, which makes our society more effective. I'd say as a person, not as Palantir. I've been a leftist since I was raised. I actually don't like war. I wanted to avoid war. I don't support war. Sometimes they are necessary, and when they are needed, America and its allies need very, very good tools. We have competitors, rivals, and enemies. They need to know our stuff works. The way to stop war is to let them know that this stuff works. Also, you know, part of the problem we have encountered in discussions in this country is that we do have an opponent on the battlefield, an enemy.
I happen to have quite a bit of respect for their culture, China, but that's what we have for them. They have a vibrant, fun, super-intelligent, large-scale culture, and they're doing a great job with AI; we have to make sure we're leaders, and we have to find ways to drag our Europe and other allies along; it's going to be hard because they also have to learn how this stuff works. And that could mean real interaction with the leaders of the sovereignty movement, rather than some fake bureaucratic stuff made up by people who don't know what they're saying; they think you can regulate this stuff like you would regulate the telecommunications industry; in fact, this is inspiring a new transformative technology that is shifting every part of society, changing our numbers and everything it touches.
operator
Thanks, Alex. As always. We have lots of personal surveys. Gotta let someone else say it. Does anyone want to say anything before we end the session? If you guys could say a few words. Tom, you want to say...
Tom
Um, I mean, you talked about deterrence very well. I think I'll just add that, of course, the goal is to deter conflict. Deterrence is not achieved by hope. You actually have to build it brick by brick, etc. Being in the trenches is about having 10 or 100 times better abilities than your opponent. If they choose to fight, they lose, so they shouldn't choose to fight.
At the other end, it comes down to the factory floor. You know, deterring conflict, as we've seen in Russia and Ukraine, is really about your productive capacity. We are at a time when we need to reinvent how we produce. We've gone through 30 or 40 years of essentially deindustrialization, and not only is it their defense industry foundation, but we're working with War Machine to partner with commercial, private, and public sector companies to achieve production using modern American manufacturing stacks.
These are two sides of the same coin. I think it boils down to, yes, freedom is driven through the trenches, but it also includes prosperity, the growth of the US economy. Behind that,
Alex Karp (CEO)
Would you like to say something? Would you like to talk about your greed? After that
Shyam Sankar (Chief Technology Officer)
I just wanted to add that, you know, the version of the sovereign AI revolution we launched last quarter, and the experience version on the ground. What we saw at the Sovereign AI Bootcamp, when we actually demonstrated rather than preached, was the customer's momentum and sense of urgency. They understood and saw it, adopted it quickly, acted with a sense of urgency because they realized how important it was, and we felt it on the ground. We were at the very beginning of it all, and we only saw it continue and expand from here.
And because I can't help it, I want to say one more thing, back to Gil's question. You know I mentioned this earlier in my review, but it's true, I don't think the labs are right in their own belief. Maybe the customers aren't fully aware that the rewards aren't there; it's the company's atomic unit not the knowledge worker. You know, you can make people more productive without generating any real return on investment.
If you measure input, it feels good to see the productivity of the person you actually are. The atomic unit is a workflow. Having a platform like AIP, you're already managing your supply chain or your S&OP or your key enterprise processes. The exhaust of these processes then allows you to go through the post-training cycle to have a model that's superior to the cutting edge in that regard. We give you the tools to gracefully integrate your people, your AI, and software to drive objective, quantifiable value for your workflow.
It's a compound interest cycle, and customers who come to our Sovereign Bootcamp really understand it. Just at our recent devcon in Miami, we had a client on their production stack as a hack week project that built a production-ready model in 48 hours that was more suited to their use case than the frontier. One person, 48 hours, that was DEVCON. So the evidence is everywhere
Dave Glazer (Chief Financial Officer)
I'd like to leave you with two financial metrics. I said in terms of net profit two years ago, until today. Our net profit has increased 8 times over the past two years, and I'm incredibly proud of that, and everything else I said, and when we look at this year, which basically guides the full year's cash flow of nearly $5 billion, I think it's something we can all be excited about.
Alex Karp (CEO)
In the end, I just, you know, many of the early investors in us were investing their own money, otherwise they would be called retail investors. Many recent people have invested in us again. You know, the Sovereign Revolution is a metaphor for your interaction with Palantir. We can help you, you can help us in your own wealth creation and our journey, and you protect us from attacks that now clearly seem absurd, stupid, and malicious, and incomprehensible, that we will never forget.
Part of the delivery here, part of what motivates me is to help ensure we fight for you and, quite frankly, make sure those who make fun of you and make fun of us are suffering. I'm so proud of it, and I'm so proud to have all of you traveling with us. Thank you all for your support.
operator
Thank you. That's today's meeting. That concludes today's conference call.