Decentralized finance (DEFI) is surprising StableCoins yields to earn Treasury yields, while smaller players in the Artificial Intelligence (AI) industry struggle to raise capital for the expansion of data centers with new GPUs.
A new StableCoin protocol called USD.AI wanted to bridge that gap By turning idle Crypto liquidity into loans for machines that train and operate artificial intelligence.
The protocol, which now counts nearly $345 million in circulation, According to a dune dashboard.
These GPUs generate revenue by selling compute time for model training and understanding, along with the cash flow that serves the debt that finances them. Lenders earn a yield from those payments rather than from token issuances, while lenders gain access to specialized financing that exceeds the risk appetite of most retail lenders.
The structure of USD.AI relies on three interlocking mechanisms designed to make real-world credit work onchain.
The first, caliber, is the legal and technical bridge between a physical GPU and its on-chain representation. Each GPU financed through the protocol is stored in an insured data center and documented under US commercial law, then tokenized as a non-negotiable token (NFT) representing a legally enforceable claim to that hardware.
Loans are issued against these tokenized receipts, allowing capital raised on-chain to fund equipment off-chain with actual collateral behind it. The next layer, the filo curator, handles the underwriting.
The curators originate and manage GPU loans while posting their own first-loss capital, meaning they absorb any initial shortfalls before lenders are affected. This structure determines the source of credit but keeps incentives aligned: curators only profit when their lenders perform.
The final component, QEV, which stands for net realizable value, governs liquidity. Instead of offering instant withdrawals, the system’s redemption requests, turning time on a market.
Users who wait are paid in installments from the borrower’s monthly payment, while those who need to get out faster can pay a premium to move the line. That premium pays the patient lenders and maintains the solvency of the loan book.
The current yield for staked Susdai ranges between 13% and 17%, supported by payments from GPU operators rather than exits or leverage loops.
Supporters of USD.AI describe it as a prototype for a broader model of “infrafi”, decentralized infrastructure finance, which could one day be extended to renewable energy projects or decentralized computing networks.
For now, its success hinges on a more immediate question: whether the GPU rental economy — a proxy for demand for AI — can remain strong enough to keep the payments flowing.
If they do, USD.AI could be Defi’s first big bridge between Onchain Capital and the real world machinery behind artificial intelligence.
