Palantir Technologies 2026 Q3 Earnings Call
Review the key takeaways and the transcript of this earnings call.
- Palantir reported third-quarter 2026 revenue of $2.313 billion, up 96% year over year and 20% sequentially, its highest reported growth rates as a public company.
- U.S. revenue was $1.937 billion, up 119% year over year and 23% sequentially, representing 84% of total revenue; U.S. commercial revenue grew 148% year over year and 29% sequentially to $984 million, surpassing U.S. government revenue for the first time.
- Commercial revenue increased 113% year over year to $1.168 billion, while government revenue increased 81% to $1.146 billion; international commercial and government revenue were $184 million and $192 million, respectively.
- Trailing 12-month revenue from the top 20 customers increased 76% year over year to $146 million per customer, and U.S. commercial customer count reached 686, up 29% year over year.
- Third-quarter total remaining deal value was $13.4 billion, up 55% year over year, and remaining performance obligations were $4.9 billion, up 88% year over year; net dollar retention was 165%.
- Adjusted operating income was $1.427 billion, a 62% margin; GAAP operating income was $1.085 billion, a 47% margin, and GAAP net income was $1.128 billion, a 49% margin.
- Adjusted free cash flow was $1.216 billion, a 53% margin, and Palantir ended the quarter with $10.4 billion in cash, cash equivalents, and U.S. Treasury securities.
STOCKNOW INSIGHTS
Continue with outlook and guidance.
Log in to unlock executive comments and Q&A highlights.
Log in for the full summaryStockNow uses AI to translate and summarize earnings calls. Accuracy and completeness are not guaranteed.
Transcript
Preview the first fifteen paragraphs, organized by speaker.
Good afternoon. I'm Ana Soro from Palantir's finance team, and I'd like to welcome you to our third quarter 2026 earnings call. We'll be discussing the results announced in our press release issued after the market closed and posted on our investor relations website. During the call, we will make statements regarding our business that may be considered forward-looking within applicable securities laws, including statements regarding our fourth quarter and fiscal 2026 results, management's expectations for our future financial and operational performance, which does not constitute formal guidance unless indicated as such in the accompanying written materials, and other statements regarding our plans, prospects, expectations, goals, and aspirations. These statements are not promises or guarantees and are subject to risks and uncertainties, which could cause them to differ materially from actual results.
Information concerning those risks is available in our earnings press release distributed after the market closed today and in our SEC filings. We undertake no obligation to update forward-looking statements except as required by law. Further, during the course of today's call, we will refer to certain adjusted financial measures. These non-GAAP financial measures should be considered in addition to, not as a substitute for or in isolation from, GAAP measures. Additional information about these non-GAAP measures, including reconciliation of non-GAAP to comparable GAAP measures, is included in our press release and investor presentation provided today. Our press release, investor presentation, and other earnings materials are available on our investor relations website at investors.palantir.com. Over the course of the call, we will refer to various growth rates when discussing our business. These rates reflect year-over-year comparisons unless otherwise stated.
Joining me on today's call are Alex Karp, Chief Executive Officer, Shyam Sankar, Chief Technology Officer, Dave Glazer, Chief Financial Officer, and Ryan Taylor, Chief Revenue Officer and Chief Legal Officer. I'll now turn it over to Ryan to start the call.
Humanity is facing world historic crossroads in how it chooses to deal with AI, and through it all, our genre-defining technology has led the way, redefining the industry. We've been shaping the AI revolution from its start, and despite being only at the very beginning of the Sovereign AI revolution, we already see how it is transforming the size of the market. We achieved revenue growth of 96% year-over-year and 20% sequentially, the highest growth rates we've ever delivered as a public company. These results are the continued financial expression of the transformational impact we're delivering for our customers, which is also evident in the depth of our customer relationships. The third quarter average trailing 12-month revenue from our top 20 customers increased 76% year-over-year to $146 million per customer. Our U.S. business is in overdrive.
We brought in $1.9 billion in U.S. revenue, which grew 119% year-over-year and 23% sequentially, representing 84% of our total revenue. This is led by our U.S. commercial business, which grew 148% year-over-year and 29% sequentially, and surpassed our U.S. government business revenue in quarter for the first time. In U.S. commercial, the average trailing 12-month revenue per customer grew 88% year-over-year and 20% sequentially. The U.S. continues to be the epicenter of the AI revolution. I won't repeat what I said last quarter, because it has become self-evident in our unparalleled results. Two years ago, we led the way with our commoditized cognition thesis, and today, the market is only beginning to recognize the full force of Sovereign AI. It's pervasive in the conversations I'm having with customers and undeniable in the results they've shared.
The COO of Optum Serve highlighted, quote, "Sovereignty doesn't mean that we have to build everything ourselves. What it does mean, though, is protecting what differentiates us while retaining the ability to evolve, to adapt, and to scale our solutions as our customers' needs change, and they change often. It means knowing where the right partner can help us move faster." L3Harris noted that by fine-tuning open-source models on its own data, a sovereign model outperformed the frontier model it had been using in less than 48 hours and cut costs by 95% in preliminary results.
NVIDIA reported that by evaluating and post-training its open models, it was able to get its smallest, most efficient model to outperform its largest model by a large gap, and can now retrain that model on just a single GPU in a couple of hours, enabling them to retrain anytime they have data to improve the decision intelligence. They stated, quote, "We captured our alpha, and we trained it into a model. We couldn't have started this process until we built the workflows in Palantir." Our customers are choosing Palantir and our Sovereign AI to transform their organizations at a furious pace. An industrial materials company piloted a nearly $3 million engagement in Q2 2026 before converting to a $29 million ACV deal last quarter.
A global athletic company converted a $1.6 million pilot that began in Q1 2026 into an approximately $21 million ACV deal last quarter. A global home appliance company converted a $1 million pilot that began in Q2 2026 into a $3 million ACV deal within the same quarter, then expanded nearly fivefold to a $15 million ACV deal in Q3 across commercial, finance, and supply chain workflows. After an AgentCamp in Q2 2026, an apparel company signed a $2.5 million pilot, then last quarter, they converted to an $11.5 million ACV deal. These commitments are downstream of the measurable impact we're delivering with our customers. With Palantir Sovereign AI, McCarthy Building Companies will save 30,000 hours annually by using post-trained Nemotron to help turn complex project requirements into contract language that allocates risk and clearly defines subcontractor obligations.
Tampa General is deploying Sovereign AI to keep sensitive data and model processing local while enforcing its own risk and privacy standards for LLM traffic and supporting the granular controls Tampa General wants to define for itself. Using fine-tuned and post-trained Nemotron to achieve this outcome, together, we also improve model performance by 7% and reduce latency by 22x. Turning to our U.S. government business, revenue grew 96% year-over-year and 18% sequentially, demonstrating continued extraordinary impact across civil and defense, with Maven as our foundational work. In Q3, the Army awarded Palantir a prime agreement to produce eight TITAN systems. Last quarter, FAA highlighted the impact of our work together, stating, quote, "That shift from scattered data to an embedded Ontology is what changes the game.
This is a modernization in operational form, changing how safety work gets done," and noted that the work they can do now would require hundreds of analysts without Palantir. As I reflect on our results, I see the furious pace continuing unabated. I am invigorated unlike ever before by the ambition I am hearing from our customers and their demand for Sovereign AI, a revolution we should remember is barely one quarter old. As extraordinary as these results are, they are only the opening chapter of Sovereign AI. I will now turn it over to Shyam.
Thanks, Ryan. Palantir is obsessed with turning tokens into economic value. It is clear that the vast majority of economically valuable problems that LLMs can solve can be solved with intelligence at or below the frontier. So what is preventing this from happening? On one end, you have the model makers who cannot conceive of solving any problem except through model improvement, so they are not even trying. On the other end, you have the hard work of diffusion, which sounds passive, but is anything but. All the value comes from the elegant integration of AI, software, and humans. This requires not a harness for the model, but something closer to a harness for the enterprise, which is exactly what AIP is. Most profoundly, the conceit that the labs have wrong is the belief that the atomic unit of the enterprise is the individual knowledge worker. It is not. The fundamental unit of the enterprise is the workflow.
You can make any set of workers more productive and generate no actual ROI. For some things, like coding, the knowledge worker is the workflow, so there is perfect product-market fit. But for the vast majority of the enterprise, it is not. Given this, it is incredibly valuable that AIP implements enterprise workflows and is able to bring AI to it in a way that generates objective, quantifiable value. For example, page 209 of Anthropic's S-1 talks about AIG's automation of insurance underwriting processes for complex commercial lines of business. That entire workflow and the elegant application of AI, software, and humans runs in AIP and was implemented by Palantir and AIG. We are the diffusion layer, not just of choice, but of necessity. There is a clear pattern emerging amongst our customers who are most successful at getting value from AI.
They have enormous ambition. Their goal is not something lame like OPEX reduction, it is market domination. The opposite is true. Customers that struggle the most are narrowly focused on reducing cost. AI is a new type of intelligence. Simply trying to shoehorn it to replacing the human misses the moment completely. The error of conflating human and machine intelligence locks you into a poorly chosen goal, one that limits your ambition and still remains out of reach. Instead, we are attacking all the work the enterprise knows it should be doing, but historically lacked the economic resources or capabilities to take on. An agent swarm watching each patient in the ICU ward to help clinicians avoid potential signs of sepsis earlier and improve patient outcomes. An agent swarm per wellhead in the Permian Basin to maximize production. An agent swarm for every customer of our customer to drive growth.
This is not about doing the same thing cheaper. It is about defining the productive work that goes beyond yesterday's frontier and capturing and compounding the new economic value it creates, and doing so with Sovereign AI. The examples are everywhere now. It is clear that post-training on your specific operational data in Ontology enables better-than-frontier performance on weights you control, protecting your alpha. Yes, it is cheaper, but the dominant variable is that more use cases can actually get to the threshold of viability for machine intelligence to generate economic value when elegantly integrated into your enterprise. Our product investments in AIP to create a closed-loop model factory are making this easy and fast. The specific post-trained model is a commodity item.
The compounding value is in the post-training pipeline that is constantly fed from your operational exhaust, captured in the Ontology, that drives automated post-training runs, model checkpoints, evaluations, and deployment across yesterday's, today's, and tomorrow's open-weight ecosystem. Our new feature of auto research and continued investment in AIP Evolve push us closer and closer to an automated loop to build, deploy, and evaluate intelligence continuously against ROI. In AIP and Sovereign AI, the Pareto optimal frontier is no longer an unachievable academic aspiration. It is a simple function call, and one we can invoke automatically on your behalf. At a recent PACOM exercise, we deployed a 30 billion parameter post-train model into space, weight, power, cooling-constrained compute, ready for the disconnected fight, with superior task-specific performance than GPT-5.6 Sol, a 3 trillion parameter model. L3Harris realized 95% cost reduction with Sovereign AI.
FULL TRANSCRIPT
Continue the full translated transcript in StockNow.
Log in to unlock every statement, the English original, and speaker-by-speaker history.
Log in for the full transcriptCall participants
8 people spoke on this call — only 2 are shown here.
PARTICIPANT LIST
View participant details in StockNow.
Log in to see executives and analysts, their roles, and complete speaking history.
Log in to view all participantsKeep exploring
