Palantir Reports 93% Revenue Growth, Pushes ‘Benchmarking’ Over AI Leaderboards

Enterprise software company Palantir reported record second-quarter 2026 results, with revenue rising 93% year-over-year to $1.935 billion, as CEO Alex Karp described the quarter as ‘otherworldly’ and pointed to accelerating demand for sovereign AI.

Its US commercial revenue grew 149% YoY to $764 million, while US government revenue increased 90% YoY to $809 million. Also, adjusted free cash flow reached $1.22 billion, representing a 63% margin.

The company raised its full-year revenue guidance to $8.15–8.16 billion. It also increased its US commercial revenue outlook to more than $3.42 billion.

Palantir has argued that enterprise AI is entering a new phase where customer-specific evaluation and sovereign model deployment matter more than frontier benchmark rankings.

CTO Shyam Sankar said during the earnings call that public AI benchmarks are no longer reliable indicators of enterprise performance, citing Palantir’s experience integrating NVIDIA’s open-weight Nemotron Ultra into its Artificial Intelligence Platform (AIP). The companies partnered in June to enable sovereign AI deployments and enterprise fine-tuning using Nemotron models. 

“We’re excited for a new era, not one of benchmaxing, but of benchmaking. The assumption that the frontier is actually the best performing is just not borne out in practice,” said Sankar. 

He revealed that within 24 hours of bringing Nemotron Ultra into Palantir’s AI stack, they found five production tasks where a standard Nemotron Ultra model without post-training beat frontier models. 

“This underscores that a handful of common benchmarks can be gamed,” added Sankar. 

Palantir also claimed its deployment approach is outperforming frontier AI providers during enterprise implementations. 

Sankar described a competitive evaluation where Palantir’s Forward Deployed Engineers (FDEs) competed directly against a frontier AI lab on an enterprise automation project. He said the frontier AI lab selected a ticketing automation use case for the evaluation but failed to produce a meaningful solution. 

Palantir instead built agent swarms for the customer's clients that generated marketing, packaging, and pricing recommendations to improve revenue and utilisation, resulting in a $10 million annual contract value deal. 

Ryan Taylor, Palantir’s Chief Revenue Officer and Chief Legal Officer, argued that enterprises relying on rented frontier models are paying rising inference costs while allowing their proprietary operational knowledge to improve these third-party foundation models.

“Enterprises that are not using Palantir are seeing their token meters spinning endlessly just to get slop without any correlation to value. This token model may be working for the labs, but it is not working for anyone else,” he said.

Taylor used the argument to position Palantir's ‘AI sovereignty’ strategy as an alternative, which enables customers to retain control over their data, business logic, actions, and security.