Ornn AI Inc. has closed a $33 million seed funding round to launch what it calls the first marketplace for AI computing power, aiming to transform compute resources into a commodity traded as transparently and efficiently as oil. The financing round was co-led by Galaxy Ventures and included participation from prominent investors like Andreessen Horowitz’s crypto fund, Nordstar, SV Angel, and previous backers such as Crucible Capital and Vine Ventures.
Founded by MIT graduates Kush Bavaria and Wayne Nelms, Ornn tackles a fragmented and opaque market where companies that need compute capacity face limited supply options, unpredictable prices, and rigid contracts without benchmarks. At the same time, data center operators struggle to manage risk and forecasting because compute contracts are negotiated in isolated deals, varying widely in price.
Ornn’s approach centers on its compute price index (OCPI), a transaction-based benchmark designed to offer a single, reliable price reference for AI computing power, similar to how commodity markets operate for oil and other physical goods. The startup partners with entities like the Intercontinental Exchange (ICE) to facilitate futures and options contracts tied directly to OCPI, enabling hedging tools that have so far been absent in the compute market.
The founders emphasize that while it took nearly a century for the oil industry to develop transparent pricing and trading mechanisms, the rapid expansion in AI infrastructure investment requires a faster evolution. They highlight compute as one of the scarcest and most valuable physical resources in the world, especially critical to AI applications. By creating standardized pricing benchmarks and a marketplace, Ornn aims to unlock capital flows toward expanding compute capacity efficiently.
This new pricing transparency and risk management capability could reshape how companies procure GPU and other computing resources, moving away from opaque contracts to a more liquid, benchmarked market. Data center operators would be able to underwrite contracts more effectively and adjust to shifting usage patterns across hardware types and regions.