DEPLOYDatabase

Company

USD.AI

usd.ai

USD.AI is an onchain lending protocol (developed by Permian Labs) that provides non-recourse, asset-backed financing secured by verified GPU infrastructure for…

Track USD.AI
Status
private

Overview

USD.AI is an onchain lending protocol (developed by Permian Labs) that provides non-recourse, asset-backed financing secured by verified GPU infrastructure for AI operators. Loans are settled on-chain against tokenized compute collateral rather than corporate balance sheets.

Verified record

Verified deployments
None on file
Active incidents
None on file

Key facts

Crucible Capital $26.8M B300 GPU loan (company-stated, Apr 2026)

Company-stated (company-stated; also Crucible): USD.AI closed a $26,821,644 non-recourse, asset-backed onchain loan to Crucible Capital secured by 72 NVIDIA B300 servers (576 GPUs total). USD.AI says 10.0% annual rate, three-year term, 70% LTV. Protocol developed by Permian Labs; framed as GPU-collateralized compute financing settled on-chain. Cap: USD.AI + Crucible April 2026 primaries; midday digest resurfaced via Meltem/trending X, not a Sep 28 announcement

HQ / positioning

Onchain lending protocol for AI infrastructure operators (GPU-backed, non-recourse credit). Developed by Permian Labs. usd.ai

$128.9M GPU financing facility (company PR)

USD.AI announced a $128.9 million GPU financing facility, which it called its largest to date, to fund a GPU cloud provider deploying NVIDIA GB200 systems in Canada (PR Newswire / USD.AI).

$100M Bullish stablecoin debt facility (company PR)

USD.AI said it secured a $100 million stablecoin debt facility from Bullish to expand non-recourse, asset-backed GPU financing for AI compute operators (PR Newswire / USD.AI).

Data & sources

Web sources

4

4 sources backing this record.View all →

Safety record

No incidents on record for USD.AI.

Only active incidents are counted. Retracted incidents are excluded from this summary but remain reachable at their canonical URLs.

Full safety record: incidents, sourcing, and exposure data →