Announcing Liquid Compute: A $15M Seed Round to Build a Regulated Venue that Commoditizes Intelligence

Today, Liquid Compute emerges from stealth with a $15 million seed round co-led by Chemistry and Firstmark. They are joined by K8 Capital, Night Capital, TrueBridge, Brainchild Holdings, UFO Holdings, Y Combinator and angel investors such as Dmitry Balyasny. Alongside the launch, the company is announcing pending applications before the CFTC for Designated Contract Market (DCM) and Derivatives Clearing Organization (DCO) status to build the first regulated exchange to trade both cash and physically settled contracts on AI infrastructure, starting with compute.

The Strategic Rationale for Regulation

Building a true financial layer for compute requires the same institutional-grade framework that governs traditional commodities. While pursuing a regulated path is long and arduous, and liquidity is difficult to bootstrap, startup entrants such as ICE in energy, CBOE in volatility, and Kalshi in event contracts prevailed and actually accrued most of the financial value in the commoditization of their multi-trillion-dollar asset classes precisely because they secured regulatory status while focusing deeply on the unique physical and structural nuances of their specific markets.

For compute, there are unique physical dynamics that are particularly challenging to overcome. Oil can be stored in a barrel. Compute cannot. If a GPU sits idle this afternoon, that capacity evaporates forever. Its value depends entirely on location, availability, configuration, and the specific workloads it can run. The heterogeneity of compute is actually far more analogous to the power markets, but instead of acting as a proxy of generating electricity, compute acts as a direct proxy for access to intelligence, making it a critical pillar of both economic and national power. The United States recognizes this reality and is pouring unprecedented capital into chips, data centers, and energy grid expansion to secure its geopolitical edge. However, building supply is only half the equation. To truly secure compute as a strategic asset, America needs robust, regulated market infrastructure to price it, allocate it efficiently, finance its rapid growth, and allow dependent companies to manage their risk.

Enabling the Compute Trader

Because intelligence is rapidly developing a measurable cost curve, massive businesses will need to manage their exposure to it. Enterprises running heavy inference will need to hedge against rising AI costs. Neoclouds will need to lock in future utilization revenues, and lenders financing data centers will need forward prices to underwrite cash flows.

This reality is driving a massive structural shift where the buyer of compute is no longer always the end user. A new class of participant is emerging, including neoclouds, brokers, and dedicated trading firms looking to buy capacity, warehouse it, resell it, arbitrage price differences, or simply take a speculative view.

Today, these participants trade on fundamentally bilateral, cumbersome infrastructure. Every resale requires a new contract, a fresh credit decision, another payment setup, and a separate SLA. As capacity changes hands, the chain of liability becomes hopelessly tangled.

Liquid Compute exists to abstract away that complexity. Before a meaningful futures market can launch, the industry must standardize exactly what is being delivered, which requires verifiable infrastructure, defined performance metrics, and credible physical reference prices. By building an Exchange for Physical (EFP) market and clearing transactions through a common framework, any participant with a balance sheet will be able to take a position on compute without needing to become a cloud operator, credit underwriter, or AI company.

This is precisely why the company has partnered with institutional market participants like Susquehanna, BGC, and Wintermute. Liquid Compute is building the infrastructure that empowers market makers, lenders, and entirely new classes of participants to trade compute as a liquid asset. More participants taking risk means a more liquid physical market, which in turn creates robust reference prices for the financial market above it. Our overall thesis is that intelligence becomes a utility supplied by many vendors, that its cost becomes a variable buyers have to actively manage, and that managing it well requires a trusted reference price, standardized contracts, and a market to trade them in.

Liquid Compute is a Y Combinator W25 company based in New York.