business

Palantir CEO Karp Attacks OpenAI, Anthropic Pricing Model

Alex Karp says runaway token costs are broken, pushing firms toward open-weight AI models and efficiency-first strategies.

Palantir CEO Alex Karp delivered a sharp rebuke of OpenAI and Anthropic on Wednesday, arguing that the dominant token-based pricing model underpinning those companies' AI services represents a fundamental failure. "Something has gone completely wrong," Karp declared, framing the current cost structure as unsustainable for enterprise customers trying to deploy AI at scale.

At the center of Karp's critique is the practice critics are calling "tokenmaxxing" — the tendency to maximize token consumption in AI workflows, which drives up costs rapidly for businesses integrating large language models into their operations. Karp contends that spiraling token expenses are creating a structural incentive problem, forcing companies to rethink how they build and deploy AI applications.

Read more Moburst Acquires Hyperzon to Expand Amazon Marketing Capabilities →

The Palantir chief argues that the economic pressure is already reshaping corporate decision-making, with a growing number of organizations pivoting toward open-weight models — AI systems whose underlying parameters are publicly available — as a cost-effective alternative to proprietary APIs from OpenAI and Anthropic. Efficiency, not raw capability, is becoming the competitive differentiator, Karp suggested.

Karp's broadside lands at a pivotal moment for the AI industry, as enterprises weigh the return on investment from generative AI adoption. Palantir, which sells AI-powered data analytics and decision-support platforms to government and commercial clients, stands to benefit if businesses sour on token-heavy cloud AI services and seek more cost-controlled deployment architectures.

The comments signal a deepening fault line in the AI sector between proprietary model providers charging per-token fees and companies advocating for leaner, more transparent alternatives. Continue reading at US Top News and Analysis.

Continue reading at US Top News and Analysis →

Frequently Asked Questions

Q.What is tokenmaxxing and why does Palantir's CEO oppose it?

Tokenmaxxing refers to the practice of maximizing token consumption in AI workflows, which rapidly drives up costs for businesses. Palantir CEO Alex Karp argues this model is fundamentally broken and unsustainable for enterprise AI deployment.

Q.What alternative to OpenAI and Anthropic does Karp recommend?

Karp suggests companies are increasingly turning to open-weight AI models — systems whose underlying parameters are publicly available — as a more cost-effective alternative to proprietary token-based APIs.

Q.Why is Palantir critical of the current AI pricing model?

Palantir contends that skyrocketing token costs force companies to prioritize spending over capability, undermining efficient AI adoption. Karp argues enterprises should focus on efficiency rather than raw model performance driven by high token consumption.

More in business →