The deal extends Palantir CEO Alex Karp’s broader push for sovereign AI and his criticism of pay-per-token pricing models used by major AI providers.
- The partnership will integrate Nebius’ AI computing and inference infrastructure directly into Palantir’s enterprise software.
- The arrangement is designed to give commercial customers greater control over their AI models, computing resources and data.
- Palantir and Nebius also plan to expand computing capacity through modular data centers, including at locations where power is already available.
Nebius Group’s (NBIS) shares rallied in early morning trade on Monday, outpeforming the broader market, after Palantir (PLTR) announced to that the Nvidia (NVDA)-backed company would be its preferred ‘soveriegn AI’ partner.
Under the partnership, Nebius’ AI computing infrastructure will run directly within Palantir’s enterprise software. The arrangement gives commercial customers another way to run their own AI models on infrastructure they control, rather than renting computing capacity from a third-party provider on a pay-per-use basis.
NBIS stock stock rose as much as 2.5% in pre-market trade, while PLTR stock edged 0.2% lower. Both were among the top trending tickers on Stocktwits at the time of writing but retail sentiment around them remained in ‘bearish’ territory.
Palantir, Nebius Push AI Compute Control
Once the integration is complete, eligible Palantir customers will be able to access Nebius’ cloud and inference infrastructure from within Palantir’s enterprise environment. That would give companies more direct control over their computing resources, data and AI models.
The companies also plan to work together to accelerate the deployment of new computing capacity, including through modular data-center deployments at locations where power is already available. The focus reflects a growing constraint in the AI infrastructure buildout: securing usable electricity and data-center capacity is becoming nearly as important as securing advanced chips.
"Nebius' compute infrastructure powers your ability to run your own AI models under conditions you control," said Palatir CEO Alex Karp. "Our ontology and their infrastructure will undergird the sovereignty our partners are demanding."
Nebius CEO Arkady Volozh echoed that sentiment, stating the partnership lets organizations "run their optimized open models on trusted infrastructure" without sacrificing performance.
Karp Takes Aim At Pay-Per-Token AI
The partnership also fits into a broader strategy Karp has been developing around what Palantir calls sovereign AI. Karp has spent much of 2026 criticizing the pay-per-token pricing model used by frontier AI companies including OpenAI and Anthropic.
In a July CNBC appearance, he called the model “insane” and said that major AI labs can charge enterprise customers large sums without delivering sufficiently clear or measurable returns.
Karp has cited examples to make that argument. Uber reportedly exhausted its entire 2026 AI coding budget by April, while another company accumulated a $500 million Anthropic bill after failing to set spending limits.
Nebius Deal Expands Palantir’s Sovereign AI Push
The Nebius agreement builds on several sovereign AI partnerships Palantir has announced this year.
In June, Palantir and Nvidia announced a sovereign AI initiative focused on running open Nemotron models on Palantir’s software for government agencies. Palantir has also partnered with Accenture to help develop sovereign AI data centers across Europe.
Read also: Samsung, TSMC Commit To ASML's Next-Gen Chip Machines, But Intel Got There First
For updates and corrections, email newsroom[at]stocktwits[dot]com.<
