Palantir Q3 2026 earnings revealed a new balance in the software company’s business: US commercial customers generated more revenue than its US government clients for the first time. Palantir reported $2.313 billion in total third-quarter revenue, up 96% from a year earlier, after the US market close on November 2, 2026. The results suggest that demand for enterprise AI systems is translating into substantial software spending, although the figures alone do not establish how much each deployment improves a customer’s operations.
The quarter ended September 30. According to Palantir’s earnings release filed with the US Securities and Exchange Commission, US commercial revenue reached $984 million, increasing 148% year over year and 29% from the previous quarter. US government revenue was $953 million, up 96% year over year. That $31 million gap is modest relative to Palantir’s total business, but it reverses a longstanding feature of a company historically associated with public-sector analytics contracts.
Commercial AI demand changes Palantir’s revenue mix
US revenue across both customer groups climbed 119% to $1.937 billion. Global quarterly revenue passed $2 billion for the first time, with growth of 20% compared with the second quarter. Palantir’s revenue figures also exceeded the roughly $2.18 billion analyst consensus cited in Reuters’ November 2 earnings coverage. This is an important distinction: the reported growth reflects revenue recognized during the quarter, rather than only new contract announcements.
Management attributes its commercial momentum to customers deploying software that connects organizational data, operational workflows and AI models. Palantir offers platforms including Foundry and its Artificial Intelligence Platform, or AIP. The practical appeal is not simply access to a chatbot. Businesses may want an AI system to draw on authorized internal information, support decisions and carry out constrained steps within existing processes.
Those requirements also create implementation challenges. A company connecting AI agents to finance, operations or customer records must decide what information each tool can access, which outputs require human approval and how to audit decisions. Related efforts across the market illustrate the same priorities: Cloudera’s enterprise AI ecosystem expansion focuses on connected data and hybrid deployments, while IBM’s AI agent identity controls address access management. These are distinct products, not evidence that the vendors deliver identical capabilities.
Bookings are substantial, but not the same as earned revenue
Palantir reported closing 212 deals worth at least $1 million during the quarter, including 107 deals of at least $5 million and 64 of at least $10 million. Total contract value booked was approximately $2.918 billion, a 6% year-over-year increase. US commercial contract value closed reached $1.463 billion, up 11% from the same period last year.
Those booking figures deserve careful interpretation. Contract value measures potential value over the life of an agreement, not cash already received or sales already recognized. Palantir’s November 2 Form 8-K incorporates financial disclosures that explain the difference, and the company cautions that some contractual options may never be exercised. Businesses can cancel some agreements under their terms. A large deal therefore supports a view of potential future demand without guaranteeing the same amount of future revenue.
The gap between 148% growth in current US commercial revenue and 11% growth in US commercial contract value closed is another number worth monitoring. It does not automatically point to weakness: timing, deal duration and previous bookings can affect quarterly comparisons. It does mean investors and customers should avoid interpreting every growth measure as interchangeable.
Profit, cash generation and a higher 2026 forecast
Palantir posted $1.085 billion in operating income under generally accepted accounting principles, representing a 47% operating margin. Net income attributable to common stockholders was about $1.128 billion, while diluted earnings per share were $0.44. Cash generated from operations totaled about $1.205 billion. These are reported accounting results, rather than projections of what later quarters will deliver.
Adjusted operating income was approximately $1.427 billion, equivalent to a 62% adjusted margin. The distinction matters because adjusted figures remove selected costs, including stock-based compensation and related payroll taxes. Such measures can help show one perspective on operating performance, but they are not substitutes for GAAP earnings and should not be treated as directly comparable across every software company.
Palantir raised its full-year 2026 revenue expectation to a range of $8.530 billion to $8.534 billion, compared with its earlier range of $8.150 billion to $8.158 billion. For the fourth quarter, management forecasts revenue between $2.649 billion and $2.653 billion. Reuters reported that the fourth-quarter outlook was above analysts’ average estimate of $2.46 billion. These numbers are company guidance, not confirmed future revenue.
What enterprise buyers should take from the results
The more useful lesson for technology buyers is about the shift from AI demonstrations to software contracts with operational scope. A platform that combines company data, workflow permissions and model output can become deeply embedded in a business. That can create value when it reduces manual work or improves decision quality, but it can also create integration expense and dependence on a particular vendor’s data model.
Before committing to a large AI deployment, customers should demand evidence tied to their own use case. Useful tests include measurable task completion rates, accuracy checks on proprietary data, clear permissions, human escalation procedures, model portability and total ongoing costs. A strong supplier earnings report is evidence of market demand; it is not independent proof that any specific customer will achieve promised savings.
Palantir also faces ordinary risks associated with rapid expansion: sustaining growth against elevated expectations, implementing complex contracts and turning bookings into recognized revenue over time. The company’s forward-looking disclosures explicitly warn that expected business outcomes can change. Competition from other enterprise software and AI providers adds to the importance of comparing capabilities rather than relying on the headline percentage alone.
For the broader enterprise AI market, the significant development is the crossover between US commercial and government revenue. The next evidence to watch is whether commercial contract growth, customer retention and realized margins continue supporting the current revenue trajectory. Palantir’s November 2 figures establish a strong quarter; they do not settle the longer-term question of how widely enterprise AI can generate durable returns.
