J.P. Morgan: Palantir growth accelerates and margins expand as JPMorgan bets on enterprise AI moving from experimentation to returns

Where is the acceleration in growth coming from?

Palantir’s second-quarter 2026 revenue was 1.935 billion dollars, up 93% year over year, compared with 85% in the first quarter. US commercial revenue grew 149%, while US government revenue rose 90%.
Globally, commercial revenue growth rose from 47% in the second quarter of 2025 to 110% in the same period of 2026, while government revenue growth climbed from 49% to 79%. JPMorgan believes companies are beginning to focus more on the return on investment from AI projects, which favors Palantir’s efforts to deepen its penetration of the commercial market.
Existing customers are also expanding their usage: second-quarter net dollar retention rose to 157% from 150% in the first quarter. The metric excludes revenue from customers acquired in the past 12 months. The company had 1049 customers in the second quarter, up 24% year over year.

Palantir’s ontology layer organizes customers’ data, business relationships, decisions and actions into a unified model. AIP then connects third-party large language models to this information, enabling AI to participate in customer workflows. JPMorgan values this ability to apply AI to actual operations.
Customers typically start with a single use case and then connect more data and business functions. As decisions and processes become embedded in the platform, moving to another platform requires this layer to be rebuilt, increasing switching costs.
The platform can adapt to different models, data management systems and computing environments, helping customers control their information and business logic. But building an ontology model itself requires substantial work, and customer concerns about vendor lock-in could also hinder new customer growth.
Frontier model providers are adding some agent orchestration capabilities to their foundation models. JPMorgan believes that if this trend progresses faster than expected, AIP’s incremental value could depend more heavily on the ontology layer and integration services led by forward-deployed engineers.

Maven has become a formally designated program of record for the US military. Data cited in the report indicate that its user base grew from about 50 thousand in January 2026 to about 100 thousand after the start of operations in Iran. JPMorgan believes the program’s status helps streamline procurement and provide multiyear funding support.
The 10 billion-dollar agreement signed by the US Army in 2025 is a ten-year framework consolidating 75 procurement pathways. More users and a broader procurement framework create opportunities for subsequent contracts, but JPMorgan also identifies converting Maven usage into new contract value as a key step.
Government procurement cycles can be long and complex. JPMorgan also notes that a major incident involving AI target identification, disputes over public-sector data use or changes in the political environment could affect business expansion.

The non-GAAP operating margin rose to 62% in the second quarter of 2026, up from 46% a year earlier. Adjusted free cash flow was 1.22 billion dollars, and the cash flow margin rose from 57% to 63%. Revenue grew faster than expenses, creating operating leverage.
Gross margin has not improved in a straight line: the second-quarter non-GAAP gross margin was 86%, above 82% a year earlier but below 88% in the first quarter. Management attributed the sequential decline to higher costs from providing cloud hosting for a government customer.
JPMorgan forecasts revenue of 8.159 billion dollars in 2026, up 82.3% year over year, with growth slowing to 52.6% in 2027. The bank initiated coverage with an Overweight rating and a December 2027 price target of 262 dollars, using a valuation of 55 times expected 2028 enterprise value to free cash flow and assuming a three-year compound annual free cash flow growth rate of 77% through 2028.